jpad journal

AND option

OR option



R.-Y. Chao1, T.-F. Chen2, Y.-L. Chang1,2,3,4


1. Department of Psychology, College of Science, National Taiwan University, Taipei, Taiwan (R.O.C.); 2. Department of Neurology, National Taiwan University Hospital, College of Medicine, National Taiwan University, Taipei, Taiwan; 3. Neurobiology and Cognitive Science Center, National Taiwan University, Taipei, Taiwan; 4. Center for Artificial Intelligence and Advanced Robotics, National Taiwan University, Taipei, Taiwan

Corresponding Author: Yu-Ling Chang, PhD (ORCID: 0000-0003-2851-3652), Department of Psychology, College of Science, National Taiwan University, No. 1, Section 4, Roosevelt Rd, Taipei 10617, Taiwan. Tel/Fax: +886-2-33663105/ +886-2-23629909; E-mail address:

J Prev Alz Dis 2020;
Published online October 28, 2020,



Background: Although evidence suggests that subjective memory complaints (SMCs) could be a risk factor for dementia, the relationship between SMCs and objective memory performance remains controversial. Old adults with or without mild cognitive impairment (MCI) may represent a highly heterogeneous group, based partly on the demonstrated variability in the level of executive function among those individuals. It is reasonable to speculate that the accuracy of the memory-monitoring ability could be affected by the level of executive function in old adults.
Objective: This study investigated the effects of executive function level on the consistency between SMCs and objective memory performance while simultaneously considering demographic and clinical variables in nondemented older adults.
Setting: Participants were recruited from both the memory clinics and local communities.
Participants: Participants comprised 65 cognitively normal (CN) older adults and 54 patients with MCI.
Measurements: Discrepancy scores between subjective memory evaluation and objective memory performance were calculated to determine the degree and directionality of the concordance between subjective and objective measures. Demographic, emotional, genetic, and clinical information as well as several executive function measurements were collected.
Results: The CN and MCI groups exhibited similar degrees of SMC; however, the patients with MCI were more likely to overestimate their objective memory ability, whereas the CN adults were more likely to underestimate their objective memory ability. The results also revealed that symptoms of depression, group membership, and the executive function level together predicted the discrepancy between the subjective and objective measures of memory function; however, the executive function level retained its unique predictive ability even after the symptoms of depression, group membership, and other factors were controlled for.
Conclusion: Although both noncognitive and cognitive factors were necessary for consideration, the level of executive function may play a unique role in understanding the equivocal relationship of the concurrence between subjective complaints and objective function measures. Through a comprehensive evaluation, high-risk individuals (i.e., CN individuals heightened self-awareness of memory changes) may possibly be identified or provided with the necessary intervention during stages at which objective cognitive impairment remains clinically unapparent.

Key words: Aging, awareness, mild cognitive impairment, memory complaints.



Subjective memory complaints (SMCs), commonly observed in older adults, refer to the self-perception of memory decline that does not require confirmation by cognitive tests. Recent studies on SMCs have revealed that they are associated with underlying brain morphometric changes (1) or increased β-amyloid deposition (2), which is consonant with dementia pathology (3).
Although evidence suggests that SMCs could be a risk factor for dementia (4, 5), the relationship between SMCs and objective memory performance remains controversial. Some studies have reported SMCs to be associated with a decline in objective memory performance (6), whereas other studies have not reported such an association (7, 8). Noncognitive factors, such as old age (9, 10), female gender (11), appearance of health conditions (e.g., hypertension and diabetes mellitus) (9, 12), low education level (9), apolipoprotein E ε4 (ApoE ε4) allele (10, 13), and depression (8-10), could also contribute to the appearance of SMCs and confound the association between SMCs and objective memory performance.
In addition to the noncognitive factors, one aspect that is crucial but has often been overlooked in studies examining the concurrence between subjective and objective memory changes is individual differences in metamemory ability. Metamemory, an aspect of high executive function level (14), is an individual’s self-awareness of his or her own memory contents and capacities. Furthermore, it is the ability of an individual to monitor or judge his or her own learning and memory efficiency. Although an age-related decline in metamemory function has been observed (15, 16), evidence suggests that aging populations, including cognitively normal (CN) older adults and individuals with mild cognitive impairment (MCI), may represent a highly heterogeneous group, based partly on the demonstrated variability in the level of executive function among those individuals (17). Accordingly, it is reasonable to speculate that the accuracy of the memory-monitoring ability could be affected by the level of executive function in both CN older adults and patients with MCI.
Thus, the aim of the present study was to examine the relationship between the level of executive function and the accuracy of SMCs while simultaneously considering noncognitive factors (including demographic variables, mood, ApoE ε4 status, and health conditions) in CN older adults and patients with MCI. To evaluate the accuracy of SMCs, we calculated the discrepancy scores between the self-reported memory concerns using a memory questionnaire and objective memory performance based on standardized neuropsychological tests, which enabled us to evaluate the accuracy of SMCs in two directions (i.e., overestimation or underestimation of objective memory performance). A lower level of executive function was hypothesized to be associated with a higher discrepancy between the subjective report of memory function and objective measurement using standardized memory tests in both CN older adults and patients with MCI, even after controlling for noncognitive (e.g., demographic variables) or clinical (e.g., health conditions and ApoE ε4 status) factors.


Materials and methods


The present study included 119 older adults, of whom 65 and 54 were classified as CN older adults and patients with MCI, respectively. Among the participants, 89 participants (48 CN and 41 MCI) were recruited from memory clinics, and 30 (17 CN and 13 MCI) were through community advertising. Individuals with any current evidence of major neurological diseases that may affect central nervous system function, psychiatric disorders, or a history of substance abuse were excluded.
Participants received a diagnosis of MCI according to the criteria recommended by the International Working Group (18). Specifically, the criteria for MCI were as follows: (1) absence of dementia, (2) defective performance on objective neuropsychological tests, and (3) generally preserved basic daily activities or the slightest impairment in instrumental activities. The objective cognitive decline was determined using the directive suggested by Jak et al. (19): the presence of at least two test scores within a cognitive domain (i.e., memory or executive function) on available neuropsychological tests (Table 1) that were one or more standard deviations less than the age-appropriate norms. Different MCI subtypes could be classified according to the aforementioned guideline. The present sample consisted of 23 patients with amnestic MCI single domain, 29 with amnestic MCI multiple domains, and 2 with nonamnestic MCI single domain. The present study was approved by the Ethics Committee and Institutional Review Board at the National Taiwan University Hospital according to the Declaration of Helsinki. Written informed consent was obtained from all participants.

Neuropsychological and Clinical Measures

A neuropsychological battery was administered to all participants. The measures included five executive function tests, namely the Matrix Reasoning and Similarities subtests of the Wechsler Adult Intelligence Scale-III (WAIS-III), category fluency test (animal and fruit), Modified Card Sorting Test (MCST), and Color Trails Test (CTT-1 and -2). A composite z-score was computed to represent each participant’s relative executive function level; the greater the positive number, the better performance it represented. Specifically, the raw score of participants’ performance on each executive function measure was first transformed into a z-score based on the norms obtained from the entire participant pool in the present study. Because lower scores (indicating that less time was required to complete the task) on the CTT reflect higher performance, the z-score of the CTT was inverted to ensure unidirectionality prior to averaging the z-scores of the five tests.
Four episodic memory tests were administered in the present study, namely the logical memory (LM) test, the visual reproduction (VR) test, the visual paired associates (VP) subtests of the Wechsler Memory Scale-III (WMS-III) (20), and the California Verbal Learning Test-II (CVLT-II) (21). A z-transforming memory composite score representing the relative performance on the episodic memory test was computed for each participant by using the method previously described; thus, a positive number represented higher memory performance. Notably, although all the four memory tests were used to classify participants’ group membership (i.e., CN versus MCI), to match the SMC subscales selected in the present study, only the two verbal episodic memory tests, namely the LM and CVLT-II, were used to compute episodic memory composite scores and for following analyses, which included both immediate (LM I, and CVLT-II List A 1-5 total recall) and delayed (LM II, and CVLT-II long-delayed free recall) recall scores.
SMCs were assessed using the Memory Complaints Inventory (22), which consisted of nine subscales designed to tap different types of reported memory problems: General Memory Problems, Verbal Memory Problems, Numeric Information Problems, Visuospatial Memory Problems, Pain Interferes with Memory, Memory Interferes with Work, Impairment of Remote Memory, Amnesia for Complex Behavior, and Amnesia for Antisocial Behavior. The first six subscales of the inventory included plausible memory complaints, and the last three subscales were intentionally designed to detect individuals with a tendency to exaggerated or feigned memory complaints. In the present study, we included scores from two subscales, namely the General Memory Problems and Verbal Memory Problems, for further analysis to sufficiently match the nature of the objective memory tests used in the present study. Lower scores on the self-evaluated questionnaire reflect a lower endorsement of memory problems by an individual. A z-transforming SMC composite score was calculated to indicate the level of endorsement of memory problems for each individual; to maintain consistency with the direction of objective memory test results, the z-scores of the SMC scores based on the questionnaire were inverted before calculating discrepancy scores. Additionally, the Framingham Stroke Risk Profile (FSRP) (23) and the Geriatric Depression Scale-Short Form (GDS-S) (24) were included to survey the participants’ cerebrovascular burden and depression status, respectively. The ApoE genotyping was conducted based on the method previously published (25), and participants were classified as ApoE ε4 carriers or non-carriers based on the appearance of at least on ε4 allele or not.

Discrepancy between Subjective and Objective Memory Evaluation

We used a modified discrepancy measure based on Miskowiak et al. (26) Specificially, the discrepancy between SMCs and objective memory performance was calculated for each participant by subtracting the standardized objective memory composite z-scores from the inversed z-transforming SMC scores. A positive value of the discrepancy score was considered to indicate that the participants’ rank ordering for their subjective evaluation was higher than their objective performance; that is, they overestimated their objective memory functioning. By contrast, a negative value of the discrepancy score was considered to indicate an underestimation of their objective memory function. Scores near zero were considered to indicate relatively high concordance between self-evaluated memory function and objective memory performance.

Statistical Analysis

Group differences were compared using analysis of variance, t-test, analysis of covariance, or chi-square tests. Statistical significance for demographic and clinical variables were set at an alpha level of 0.05, whereas the significance level for neuropsychological measures was set at p < 0.003 based on Bonferroni correction to avoid inflated type I errors. The discrepancy scores were checked for normal distribution using the Kolmogorov–Smirnov test, and the result indicated that it did not violate the null hypothesis (p > 0.20, with a mean score of 0.01, standard deviation [SD] = 1.53).
Hierarchical regression analyses were conducted to examine the predictive ability of the level of executive function for determining the discrepancy between subjective and objective memory evaluations; the corresponding alpha level was set at 0.05. Specifically, demographic variables including age, sex, and education were considered simultaneously in the first step. Subsequently, clinical variables, including FSRP, depressive state, ApoE ε4 status, and group membership (i.e., CN versus MCI), were considered in the second step. Finally, the composite z-score of executive function level was considered in the third step. All statistical analyses were conducted using SPSS (version 22.0. IBM Corp, Armonk, NY, USA).



Demographics, Clinical Data, and Neuropsychological Performance

The two groups differed in age (F(1, 117) = 10.49, p = 0.002), education (F(1, 117) = 9,32, p = 0.003), and FSRP (F(1, 117) = 6.02, p = 0.016,Table 1), but they did not differ in the distribution of sex (χ2(2, N = 119) = 0.44, p > 0.05), frequency of ApoE ε4 carriers (χ2(2, N = 119) = 0.20, p > 0.05), scores on the depression measures (F(1, 117) = 1.39, p > 0.05), or distribution of recruitment source by the diagnostic group (χ2(2, N = 119) = 0.07, p > 0.05).

Table 1. Demographic, clinical, and cognitive characteristics with means (SDs) in the groups comprising cognitively normal older adults and patients with mild cognitive impairment

Abbreviations: CN, cognitively normal; CVLT-II, California Verbal Learning Test; FSRP, the Framingham Stroke Risk Profile; GDS-S, Geriatric Depression Scale-Short Form; GMCI, Green’s Memory Complaints Inventory; LM, Logical Memory; MCI, mild cognitive impairment; MCST, Modified Card Sorting Test; EF z-score, executive function composite z-score; VP, Visual Paired Associate; VR, Visual Reproduction Associate; η2, effect size of analysis of variance or analysis of covariance; SD, standard deviation; * p < 0.05. ** p < 0.003 (Bonferroni correction); †Group difference controlling for age, education, and FSRP. ‡ Time difference was calculated by subtracting CTT-1 from CTT-2.


After the effects of age, education, and FSRP were controlled for, the performance of the CN group was higher than that of the MCI group on all executive function measures (see Table 1), including the WAIS-III Matrix Reasoning subtest (F(1, 113) = 18.19, p < 0.001), VF (F(1, 113) = 14.08, p < 0.001), MCST (F(1, 114) = 26.40, p < 0.001), and executive function composite z-score (F(1, 114) = 30.68, p < 0.001), except for the WAIS-III Similarities subtest (F(1, 113) = 6.60, p > 0.003) and CTT measure (i.e., CTT-2 − CTT-1; F(1, 114) = 6.08, p > 0.003). Furthermore, the performance of the CN group was higher than that of the MCI group on all episodic memory measures, including the immediate recall (F(1, 114) = 33.31, p < 0.001), delayed recall (F(1, 114) = 55.19, p < 0.001), and delayed recognition (F(1, 114) = 43.51, p < 0.001) of the WMS-III VR subtests; immediate (F(1, 114) = 39.61, p < 0.001) and delayed recall (F(1, 114) = 26.01, p < 0.001) of the VP subtests; immediate (F(1, 114) = 44.21, p < 0.001) and delayed recall (F(1, 114) = 57.90, p < 0.001) of the LM subtest; immediate List A 1-5 total recall (F(1, 112) = 99.68, p < 0.001) and long-delayed free recall (F(1, 112) = 106.55, p < 0.001) of the CVLT-II; and episodic memory composite z-score (F(1, 114) = 105.36, p < 0.001).

Subjective and Objective Memory Discrepancy Measures

No differences in the SMC scores were observed between the groups (F(1, 114) = 1.12, p > 0.05) after age, education, and FSRP were controlled for. In addition, no differences in the SMC scores were observed by recruitment source of participants in the CN group (T(63) = -1.03, p > 0.05) or in the MCI group (T(52) = -1.38, p > 0.05). However, the absolute discrepancy score values differed between the two groups (F(1,114) = 14.60, p < 0.001, eta square = 0.11) after the effects of age, education, and FHS-stroke risk were controlled for; in which the CN group exhibited a relatively higher accuracy (i.e., values trended toward zero regardless of the directionality) than the MCI group in estimating objective memory. Furthermore, the two groups demonstrated significant differences in discrepancy scores (F(1, 114) = 16.71, p < 0.001) when directionality (i.e., overestimation versus underestimation) was considered and age, education, and FSRP were controlled for. We observed that this differential pattern of discrepancy scores remained significant even after further controlling for the level of executive function (F(1, 113) = 7.50, p = 0.007, η2 = 0.06); this finding suggests that the CN group, despite its relatively high objective memory performance, tended to endorse more memory complaints than the MCI group did, but the MCI group exhibited an opposite pattern.
We also compared the frequencies of overestimation and underestimation of memory function between the two groups by dichotomizing the discrepancy scores (i.e., ≥0 and <0). The results showed that the two groups exhibited significant differences in the frequency distribution of the two discrepancy categories (χ2(2, N = 119) = 19.13, p < 0.001). The number of participants underestimating their objective memory ability was higher in the CN group than in the MCI group, whereas that of participants overestimating their objective memory ability was higher in the MCI group than in the CN group (Figure 1). We further analyzed the demographic and clinical characteristics of the two subgroups (i.e., underestimation versus overestimation of objective memory ability) in each of the CN and MCI groups. Within the CN group, the underestimation and the overestimation subgroups did not differ in age (T(63) = -1.16, p = > 0.05), education(T(63) = 0.43, p = > 0.05), distribution of sex (χ2(1, N = 65) = 1.87, p > 0.05), FSRP (T(63) = -1.03, p = > 0.05), frequency of ApoE ε4 carriers (χ2(1, N = 65) = 3.11, p > 0.05), depression status(T(63) = 0.83, p > 0.05), or executive function (T(63) = 1.61, p > 0.05). Notably, a trend of higher executive function was observed among CN participants who underestimated their memory ability (executive function z-score = 0.44 ± 0.53) compared with those who overestimated their memory ability (executive function z-score = 0.22 ± 0.54). Similarly, within the MCI group, the underestimation and the overestimation subgroups did not differ in age (T(52) = -1.90, p = > 0.05), education (T(52) = 0.58, p = > 0.05), distribution of sex (χ2(1, N =54) = 0.02, p > 0.05), frequency of ApoE ε4 carriers (χ2(1, N =54) = 0.89, p > 0.05), or depression status(T(52) = 0.19, p = > 0.05), but the subgroup with overestimation of objective memory ability demonstrated a higher FSRP score(T(52) = -2.43, p = 0.001, FSRP score = 16.6% ±12.15) and marginally lower executive function (T(52) = 1.87, p = 0.06, executive function z-score = -0.54 ± 0.70) compared to the underestimation subgroup (FSRP score = 8.5% ± 4.38; executive function z-score = -0.16 ± 0.57).

Figure 1. Pie charts depicting a comparison of frequency distribution of underestimation (negative discrepancy z-scores) and overestimation (positive discrepancy z-scores) of objective memory function between the groups comprising cognitively normal older adults and patients with mild cognitive impairment


Notably, when participants were classified into small (i.e., z-scores within the range between +1 to −1) versus large discrepancy scores (z-scores > + 1 or < −1) without considering the directionality of the scores, significantly more MCI patients (55.6%) obtained large discrepancy scores, indicating a larger misjudgment for their memory ability compared than for the CN group (26.2%) (χ2(1, N = 119) = 10.67, p = 0.001). We further analyzed the demographic and clinical characteristics of the two subgroups (i.e., small versus large discrepancy) in each of the CN and MCI groups. Within the CN group, the two subgroups did not differ in age (T(63) = 0.24, p = > 0.05), education(T(63) = -0.99, p = > 0.05), distribution of sex (χ2(1, N = 65) = 0.80, p > 0.05), FSRP (T(63) = -0.68, p = > 0.05), frequency of ApoE ε4 carriers (χ2(1, N = 65) = 1.42, p > 0.05), or depression score (T(63) = -0.97, p > 0.05). The two subgroups did not differ in executive function (T(63) = −0.96, p > 0.05). However, a trend toward higher executive function was observed among those with a greater degree of misjudgment (executive function z-score = 0.48 ± 0.51) compared with patients with mild misjudgment (executive function z-score = 0.33 ± 0.55). Within the MCI group, the two subgroups did not differ in age (T(52) = -1.22, p = > 0.05), education (T(52) = -0.72, p = > 0.05), distribution of sex (χ2(1, N =54) = 0.84, p > 0.05), FSRP score (T(52) = -0.31, p > 0.05), frequency of ApoE ε4 carriers (χ2(1, N =54) = 0.73, p > 0.05), depression score (T(52) = 1.37, p = > 0.05), or executive function (T(52) = 0.17, p > 0.05).

Relationships between Executive Function Level and Discrepancy Score

Hierarchical multiple regression analysis (Table 2) demonstrated that increased endorsement of depressive symptoms predicted negative discrepancy scores (i.e., increase in self-reporting of memory concerns and an underestimation of objective memory ability) (β = −0.17, p = 0.046), and diagnosis of MCI predicted positive discrepancy scores (i.e., overestimation of objective memory ability) (β = 0.26, p = 0.008). Moreover, the level of executive function (β = −0.24, ΔR2 = 0.03, p = 0.027) explained the unique variances in the discrepancy scores in addition to the demographic and clinical variables; a higher level of executive function was associated with underestimation of objective memory ability (Figure 2).

Table 2. Hierarchical regression models with predictive ability of demographic, clinical, and executive function level for discrepancy between subjective and objective memory evaluations

Abbreviations: FSRP, the Framingham Stroke Risk Profile; GDS-S, Geriatric Depression Scale-Short Form; Group, participants were classified as cognitively normal older adults or patients with mild cognitive impairment; ApoE ε4: participants were classified as ApoEε4 carriers or noncarriers; EF z-score, executive function composite z-score. * p < 0.05; ** p < 0.01.

Figure 2. Scatter plot of the relationship between executive function level and discrepancy scores between subjective memory complaints and objective memory performance. A low level of executive function was associated with positive discrepancy scores (i.e., overestimation of objective memory functioning).* p < 0.05; ** p < 0.01



The primary objective of this study was to investigate the effect of the level of executive function on the consistency between SMCs and performance on objective memory function measures while considering demographic (e.g., age, education, and sex), emotional (i.e., symptoms of depression), and clinical (e.g., ApoEε4 status, and health conditions related to cardiovascular risk factors) variables in a sample comprising CN older adults and patients with MCI. An analysis of the discrepancy scores between the subjective and objective measures revealed that although the symptoms of depression, group membership, and level of executive function together predicted the discrepancy between the subjective and objective measures of memory performance, the level of executive function retained its predictive ability even after the symptoms of depression, group membership, or other factors were controlled for.
In this study, we used five executive function measures to assess the relationships between the level of executive function and the consistency between the subjective and objective measures of memory functioning; these five measures were thought to involve prefrontal function (27) and were essential for successful self-monitoring (28), such as reasoning, ability to use external feedback to modify thinking or behavior, and shifting and updating information. As predicted, we found that a lower level of executive function was associated with a greater degree of discrepancy between subjective and objective measures of memory function in our sample of elderly participants. This result is consistent with the emerging literature that has demonstrated a relationship between reduced awareness of memory loss and frontal lobe dysfunction in patients with Alzheimer disease (29) and MCI (30).
Another critical finding in this study was that the CN group generally had higher accuracy in estimating their memory ability compared with the MCI group, despite the CN and MCI groups exhibiting similar degrees of memory complaints. Furthermore, group membership predicted the discrepancy scores. Notably, the two groups exhibited different patterns of discrepancy scores: The participants in the CN group were more likely to underestimate their objective memory ability. However, 63% (27 out of 43) of participants in the underestimation subgroup underestimated their memory ability within a relatively mild range (i.e., discrepancy z-scores > −1). By contrast, the participants in the MCI group were more likely to overestimate their objective memory ability. Such findings appear to be consistent with those of a recent study by Fragkiadaki et al. (31), who used an “in-session” cognitive efficiency measure and found that CN older adults underestimated their performance, whereas patients with MCI overestimated their performance on a task. This inclination to underestimate actual performance was also reported by Vannini et al. (30) in CN older people with β-amyloid deposition; the authors introduced the term “hypernosognosia” to indicate that heightened memory self-awareness may be the first stage of progression toward Alzheimer disease in a hypothetical memory awareness model. Notably, the subset of CN participants in the present study with a larger degree of underestimation (i.e., discrepancy z-scores < −1) of their objective memory ability exhibited a trend of higher executive function compared with participants with mild underestimation of their memory ability. Despite its counterintuitive nature, our findings may suggest that individuals with hypernosognosia may have relatively high self-monitoring ability and may have experienced some memory loss relative to a previously higher baseline memory function, which could not be captured by the neuropsychological tests employed here. Although we attempted to include multiple measures of memory in the present study to ensure the reliability and sensitivity of the measurements, these measures might still not be sufficiently sensitive to detect subtle within-person memory declines, because the calculations of the objective memory index are completely based on comparisons with group norms. Consequently, the CN older adults with hypernosognosia may represent a group of people who are at risk of dementia in the future, particularly when factors such as symptoms of depression or healthy conditions, which could potentially confound the interpretations of their “worries,” were considered. Following up CN older adults with hypernosognosia longitudinally to further examine such a hypothesis is crucial.
Consistent with accumulating studies that have regarded depression as a crucial factor accounting for SMCs (32), in the present study, we also observed that an increase in the symptoms of depression was predictive of an increase in self-reporting of memory concerns and an underestimation of memory ability. Previous studies have indicated that a higher frequency of SMCs were strongly associated with a higher number of depressive symptoms, regardless of the objective cognitive performance (8). By contrast, other evidence indicates late-life depression to be a risk factor for progression to dementia (32). Although the exact relationships among depressive symptoms, SMC, and objective cognitive function warrant further investigation, the present study extends previous findings by demonstrating that the contribution of the level of executive function is crucial because its contribution to the concurrence between subjective and objective measures is unique and is separate from that of the symptoms of depression per se.
Despite the potential clinical value of our findings, our study has limitations. First, we included only depressive symptoms in the analyses, and we did not consider other affective (e.g., anxiety) or personality factors, which might also affect the concurrence between the objective and subjective memory measures as reported by previous studies (32, 33). Second, the cross-sectional design of the present study precluded us from investigating the relationship among the level of executive function, discrepancy scores, and subsequent function declines. This design also limited our ability to examine the linear continuum versus nonlinear evolution from SCD to MCI. In addition, the sample size of the current study was relatively small, particularly when considering the number of predictive variables used for the regression model. The small sample size also prevented us from clarifying the heterogeneity among the older adults, particularly among the patients with MCI. Despite the aforementioned limitations, the present study is the first to consider various factors in investigating the directionality of the concurrence between subjective and objective memory function. We also extend previous studies by using a more ecologically relevant self-report measure to survey the individuals’ subjective memory concerns and objectively measure memory function through standardized cognitive tests.
In conclusion, the present study reveals the complexity involved in understanding the meaning of SMCs. Although both noncognitive and cognitive factors were necessary for consideration, the level of executive function may play a unique role in understanding the equivocal relationship of the concurrence between subjective complaints and objective function measures in the literature. Through a comprehensive evaluation, high-risk individuals (i.e., CN individuals with hypernosognosia) may possibly be identified or provided with the necessary intervention during stages at which objective cognitive impairment remains clinically unapparent. Inclusion of biomarkers and longitudinal follow-up data can provide additional information on the neural mechanism underlying the discordance.


Funding: This work was supported by the Taiwan Ministry of Science and Technology (grant numbers 107-2314-B-182A-065 and108-2410-H-002-106-MY2 to Y.L.C.). The sponsors had no role in the design and conduct of the study; in the collection, analysis, and interpretation of data; in the preparation of the manuscript; or in the review or approval of the manuscript.

Acknowledgements: The authors thank Yen-Shiang Chiu, Chia-Hua Lin, and Yi-Yuan Zhuo for assistance in data collection, and Dr. Jung-Lung Hsu for constructive feedbacks.

Conflict of interest: The authors have no conflict of interest to report.

Ethical Standards: The institutional Review Board(IRB) approved this study, and all participants gave informed consent before participating.



1. Peters F, Villeneuve S, Belleville S. Predicting progression to dementia in elderly subjects with mild cognitive impairment using both cognitive and neuroimaging predictors. J Alzheimers Dis 2014;38(2): 307-18..
2. Wolfsgruber S, Polcher A, Koppara A, Kleineidam L, Frolich L, Peters O, et al. Cerebrospinal Fluid Biomarkers and Clinical Progression in Patients with Subjective Cognitive Decline and Mild Cognitive Impairment. J Alzheimers Dis, 2017.
3. Jessen F, Wolfsgruber S, Wiese B, Bickel H, Mosch E, Kaduszkiewicz H, et al. AD dementia risk in late MCI, in early MCI, and in subjective memory impairment. Alzheimers Dement 2014;10(1): 76-83.
4. Buckley RF, Maruff P, Ames D, Bourgeat P, Martins RN, Masters CL, et al. Subjective memory decline predicts greater rates of clinical progression in preclinical Alzheimer’s disease. Alzheimers Dement 2016;12(7): 796-804.
5. Abner EL, Kryscio RJ, Caban-Holt AM, Schmitt FA. Baseline subjective memory complaints associate with increased risk of incident dementia: the PREADVISE trial. J Prev Alzheimers Dis 2015;2(1): 11-16.
6. Seo EH, Kim H, Lee KH, Choo IH. Altered Executive Function in Pre-Mild Cognitive Impairment. J Alzheimers Dis 2016;54(3): 933-40.
7. Edmonds EC, Delano-Wood L, Galasko DR, Salmon DP, Bondi MW, Alzheimer’s Disease Neuroimaging I. Subjective cognitive complaints contribute to misdiagnosis of mild cognitive impairment. Journal of the International Neuropsychological Society : JINS 2014;20(8): 836-47.
8. Zlatar ZZ, Muniz M, Galasko D, Salmon DP. Subjective Cognitive Decline Correlates With Depression Symptoms and Not With Concurrent Objective Cognition in a Clinic-Based Sample of Older Adults. The journals of gerontology Series B, Psychological sciences and social sciences, 2017.
9. Chen ST, Siddarth P, Ercoli LM, Merrill DA, Torres-Gil F, Small GW. Modifiable risk factors for Alzheimer disease and subjective memory impairment across age groups. PloS one 2014;9(6): e98630.
10. Chhetri JK, de Souto Barreto P, Cantet C, Cesari M, Coley N, Andrieu S, et al. Trajectory of the MAPT-PACC-Preclinical Alzheimer Cognitive Composite in the Placebo Group of a Randomized Control Trial: Results from the MAPT Study: Lessons for Further Trials. J Prev Alzheimers Dis 2018;5(1): 31-35.
11. Jonker C, Geerlings M, Schmand B. Are Memory Complaints Predictive for Dementia? A Review of Clinical and Population-based Studies. Int J Geriatr Psychiatry 2000;15: 983-91.
12. Zhang X, Schmitt FA, Caban-Holt AM, Ding X, Kryscio RJ, Abner E. Diabetes mitigates the role of memory complaint in predicting dementia risk: Results from the Prevention of Alzheimer’s Disease with Vitamin E and Selenium Study. J Prev Alzheimers Dis 2017;4(3): 143-48.
13. Krell-Roesch J, Woodruff BK, Acosta JI, Locke DE, Hentz JG, Stonnington CM, et al. APOE epsilon4 Genotype and the Risk for Subjective Cognitive Impairment in Elderly Persons. J Neuropsychiatry Clin Neurosci 2015;27(4): 322-5.
14. Royall DR, Lauterbach EC, Cummings JL, Reeve A, Rummans TA, Kaufer DI, et al. Executive Control Function: A Review of Its Promise and Challenges for Clinical Research. A report from the Committee on Research of the American Neuropsychiatric Association. The Journal of neuropsychiatry and clinical neurosciences 2002;14(4): 377-405.
15. Morson SM, Moulin CJ, Souchay C. Selective deficits in episodic feeling of knowing in ageing: a novel use of the general knowledge task. Acta Psychol (Amst) 2015;157: 85-92
16. Sacher M, Isingrini M, Taconnat L. Effects of aging and divided attention on episodic feeling-of-knowing accuracy. Acta Psychol (Amst) 2013;144(2): 258-63.
17. Chang YL, Jacobson MW, Fennema-Notestine C, Hagler DJ, Jr., Jennings RG, Dale AM, et al. Level of executive function influences verbal memory in amnestic mild cognitive impairment and predicts prefrontal and posterior cingulate thickness. Cereb Cortex 2010;20(6): 1305-13.
18. Winblad B, Palmer K, Kivipelto M, Jelic V, Fratiglioni L, Wahlund LO, et al. Mild cognitive impairment–beyond controversies, towards a consensus: report of the International Working Group on Mild Cognitive Impairment. Journal of Internal Medicine 2004;256(3): 240-46.
19. Jak AJ, Bondi MW, Delano-Wood L, Wierenga C, Corey-Bloom J, Salmon DP, et al. Quantification of five neuropsychological approaches to defining mild cognitive impairment. Am J Geriatr Psychiatry 2009;17(5): 368-75.
20. Hua MSC, B.S.; Lin, K.N.; Yang, J.M.; Lu, S.R.; Chen, H.Y. Wechsler Memory Scale (WMS), Third Edition. Taipei, Taiwan: The Chinese Behavioral Science Corporation.
21. Delis DCK, J. H.; Kaplan, E.; Ober, B. A. Manual for the California Verbal Learning Test, (CVLT-II). San Antonio, TX: The Psychological Corporation, 2000.
22. Green P, Gervais R, Merten T. Das Memory Complaints Inventory (MCI): Gedächtnisstörungen, Beschwerdenschilderung und Leistungsmotivation. Neurologie & Rehabilitation 2005;11(3).
23. D’Agostino RB, Wolf PA, Belanger AJ, Kannel WB. Stroke risk profile: adjustment for antihypertensive medication. The Framingham Study. Stroke 1994;25(1): 40-3.
24. Liu CY, Wang SJ, Teng EL, Fuh JL, Lin CC, Lin KN, et al. Depressive disorders among older residents in a Chinese rural community. Psychological Medicine 1997;27: 943-49.
25. Chang YL, Yen YS, Chen TF, Yan SH, Tseng WY. Clinical Dementia Rating Scale Detects White Matter Changes in Older Adults at Risk for Alzheimer’s Disease. Journal of Alzheimer’s disease : JAD 2016;50(2): 411-23.
26. Miskowiak KW, Petersen JZ, Ott CV, Knorr U, Kessing LV, Gallagher P, et al. Predictors of the discrepancy between objective and subjective cognition in bipolar disorder: a novel methodology. Acta Psychiatr Scand 2016;134(6): 511-21.
27. Stuss DT, Levine B. Adult clinical neuropsychology: lessons from studies of the frontal lobes. Annual review of psychology 2002;53(1): 401-33.
28. Galeone F, Pappalardo S, Chieffi S, Iavarone A, Carlomagno S. Anosognosia for memory deficit in amnestic mild cognitive impairment and Alzheimer’s disease. Int J Geriatr Psychiatry 2011;26(7): 695-701.
29. Zamboni G, Drazich E, McCulloch E, Filippini N, Mackay CE, Jenkinson M, et al. Neuroanatomy of impaired self-awareness in Alzheimer’s disease and mild cognitive impairment. Cortex 2013;49(3): 668-78.
30. Vannini P, Amariglio R, Hanseeuw B, Johnson KA, McLaren DG, Chhatwal J, et al. Memory self-awareness in the preclinical and prodromal stages of Alzheimer’s disease. Neuropsychologia 2017;99: 343-49.
31. Fragkiadaki S, Kontaxopoulou D, Beratis IN, Andronas N, Economou A, Yannis G, et al. Self-awareness of cognitive efficiency: Differences between healthy elderly and patients with mild cognitive impairment (MCI). J Clin Exp Neuropsychol 2016;38(10): 1144-57.
32. Rabin LA, Smart CM, Amariglio RE. Subjective Cognitive Decline in Preclinical Alzheimer’s Disease. Annu Rev Clin Psychol 2017;13: 369-96.
33. Balash Y, Mordechovich M, Shabtai H, Giladi N, Gurevich T, Korczyn AD. Subjective memory complaints in elders: depression, anxiety, or cognitive decline? Acta Neurol Scand 2013;127(5): 344-50.



F. Buckinx1,2, M. Aubertin-Leheudre1,2


1. Department of Exercise Sciences, Université du Québec à Montréal (UQAM), Montréal (Qc), Canada; 2. Centre de Recherche de l’Institut Universitaire de Gériatrie de Montréal (CRIUGM), Montréal (Qc), Canada

Corresponding Author: Aubertin-Leheudre Mylene, Département des Sciences de l’activité physique, Faculté des Sciences, UQAM, Pavillon Sciences Biologiques, SB-4615, 141, Avenue du Président Kennedy, Montréal, Québec, Canada, H2X 1Y4. Email:

J Prev Alz Dis 2021;1(8):110-116
Published online August 7, 2020,



Aging is associated with cognitive declines leading to mild cognitive impairments or Alzheimer disease. Nutrition appear to protect from aging. Some dietary factors could either increase or protect against cognitive declines. This article aimed to provide GRADE recommendations related to nutrition aspects able to prevent or to treat cognitive impairments. A comprehensive literature review was performed using Medline database. The GRADE approach was used to classify quality of the existing evidence (systematic review or meta-analysis).The GRADE process led us to formulate seven key nutritional recommendations to manage cognitive declines, but did not allow us to do it for protein, vitamin B or antioxidants. Thus, 1) adherence to a Mediterranean diet (GRADE 1B); 2) high-level of consumption of mono- or poly- unsaturated fatty acids combined to a low consumption of saturated fatty acids (GRADE 1B); 3) high consumption of fruits and vegetables (GRADE 1B); 4) higher vitamin D intake (GRADE 1C) than the recommended daily allowance. In addition, a ketogenic diet, a low consumption of whole-fat dairy products or a caloric restriction are promising nutritional habits although the evidence does not yet support widespread uptake (GRADE 2C). In conclusion, nutrition is an important modifiable factor to prevent or protect against cognitive decline. Nevertheless, more studies are required to determine specific guidelines such as duration and amounts of nutrients to help older adult to maintain a healthy cognitive life.

Key words: GRADE process, aging, diet guidelines, cognitive impairment.



Cognitive impairment is a public health problem due to its increasing prevalence in the aging population (1). It is estimated that 24.3 million people are affected by dementia worldwide, with 4.6 million new cases annually (2). Despite pharmacological advances, there are not yet effective treatments to delay or reverse cognitive impairment (3). Moreover, there is limited knowledge of Alzheimer disease (AD) modifiable risk factors. One of these factors is nutrition (4). According to the World Health Organization (WHO), good nutrition – an adequate, well-balanced diet – is a cornerstone of good health. Immunity system, disease, physical and mental development can be negatively impacted by poor nutrition (5). Therefore, the role of nutrition in cognitive health outcomes has been studied. More specifically, nutrition has been investigated in relation to its ability to maintain cognitive function (3). Caloric intake and diet composition seem to have large and lasting effects on cognition (6). There is more and more evidence suggesting the protective role of certain dietary components (e.g., antioxidant or vitamin nutrients, fish, dietary fats) in the risk of age-related cognitive decline and AD. Several systematic reviews of the literature set out to summarize the evidence considering diet as a protective or risk factor for cognitive decline (7-9); however, clear recommendations have not yet been established. In view of these promising data, consensus is needed on the best eating habits to prevent or to treat cognitive impairment. In this sense, the GRADE (Grades of Recommendation, Assessment, Development, and Evaluation) approach guide about the quality of underlying evidences by providing grade strength of recommendations in health field. The GRADE system classifies evidence quality according to factors such as study methodology, consistency and precision of the results or directness of the finding (10). Therefore, this article aimed to grade, classify and provide recommendations for the preferred diet to prevent or to treat cognitive impairment.



Study design

This literature review aimed to provide recommendations using GRADE classification on the best nutritional habits to prevent or to treat cognitive impairment, through a narrative synthesis.

Literature search

The following PICOS strategy was defined: Population or disease: Humans; Intervention: nutrition; Comparator: nothing; Outcome: cognitive function; Study design: systematic reviews or meta-analysis.
A comprehensive literature review was performed using the Medline (via Ovid) database. The search was limited to published systematic reviews and meta-analyses and only those published in English were eligible. The search strategy combined keywords and Medical Subject Heading (MeSH) terms concerning the population (e.g. humans, men and women), the outcomes (e.g. cognitive impairment, cognitive decline, Alzheimer disease), the intervention (e.g. nutrition, diet, eating habits) and the type of study (e.g. systematic review, meta-analysis) by using Boolean indicators. Only studies published in the last ten years were eligible.
Data extraction and analysis
The GRADE (Grades of Recommendation, Assessment, Development, and Evaluation) approach classifies quality of the existing evidence available in systematic reviews and meta-analysis. Evidence was graded according to the classification of the American Academy of Sleep Medicine (AASM) and adapted from the Oxford Centre for Evidence-based Medicine Levels of Evidence, as presented in Table 1 (11).
Based on the GRADE process, a consensus statement was proposed for the best nutritional habits to prevent or to treat cognitive impairment.

Table 1. Classification of Evidence



Based on the literature, the GRADE process led us to formulate seven key nutritional recommendations (Table 2). The first is global and concerns dietary pattern/habits while the fifth recommendation is a specific diet. The other recommendations focus on specific nutrients. However, based on lack of evidence, any other recommendation about the effect of other nutrients on cognition couldn’t be realized. Our seven recommendations are developed below.

Table 2. Summary of Grading and recommendations for the preferred diet to prevent or to treat cognitive impairment


1) We recommend adherence to a Mediterranean diet to decrease the risk of cognitive decline: GRADE 1B

A Mediterranean diet is characterized by high intake of vegetables, fruits and nuts, cereals, fish and monounsaturated fatty acids (MUFA); relatively low intakes of meat and dairy products; and moderate consumption of alcohol.
This first recommendation is based on five recent systematic reviews published between 2011 and 2019 (7-9, 12, 13). Yusufov et al. (2017) found that 50/64 studies revealed an association between diet and Alzheimer disease (AD) incidence (7). More specifically, there was a 34% to 60.7% reduction in odds of AD with increase in adherence to a Mediterranean diet (7). These results are in line with those of Samadi et al. (2019) showing that 12/26 studies reported that Mediterranean diet was associated with the decrease of AD risk. The review of Aridi et al. (2017) also corroborates these results, showing that Mediterranean diet could protect against cognitive decline and decrease the risk of AD (10). Then, Otaegui-Arrazola et al. (2013) highlighted that 5/6 studies report an association between higher adherence to Mediterranean diet and decreased AD risk. Interestingly, in 2/3 studies, the risk of MCI and MCI conversion to AD was found to be lower in subjects with better adherence to this dietary pattern. In this review, Mediterranean diet has been demonstrated to improve cognitive functions (12). Finally, as reported by Solfrizzi et al. (2011), a high adherence to Mediterranean diet in AD patients could slow cognitive decline, reduced risk of progression from Mild Cognitive impairment (MCI) to AD, reduced risk of AD and decreased all-cause mortality (8).
In conclusion, these reviews join to say that adherence to a Mediterranean diet may help to decrease the incidence of AD.

2) We recommend a high level of consumption of mono- and polyunsaturated fatty acids and a low consumption of saturated fatty acids, to reduce the risk of cognitive decline: GRADE 1B

Fish and fish oil are rich sources of omega-3 fatty acids (i.e. monounsaturated fatty acids), specifically eicosa-pentaenoic acid (EPA) and docosa-hexaenoic acid (DHA). Alpha-linolenic acid (ALA) is an omega-3 fatty acid present in seeds and oils, green leafy vegetables, and nuts and beans. Linoleic acid (LA), an omega-6 fatty acid (i.e. polyunsaturated fatty acids), is present in grains, meats, and the seeds of most plants. Saturated fats are present in large quantities in meat, processed meat, milk, yogurt, cheese, butter are predominantly palmitic acid, predominantly in the form of palmitic acid.
This recommendation is based on four systematic reviews, including two meta-analysis, published between 2011 and 2019 (8, 12, 14, 15). The meta-analysis conducted on nine studies by Cao et al. (2019) highlighted that the highest category of saturated fat intake was associated with an increased risk of cognitive impairment (RR = 1.40; 95% CI: 1.02-1.91) and AD (RR: 1.87, 95% CI: 1.09-3.20) (14). However, the total and unsaturated fat intake was not statistically significantly associated with cognitive outcomes (15). The study of Solfrizzi et al. published in 2001 showed that elevated dietary monounsaturated fatty acids and n-3 polyunsaturated fatty acids and high fish consumption may have a beneficial effect on the risk of dementia (8). In 2013, Otaegui-Arrazola et al. have performed a systematic review concluding that the consumption of n-3 polyunsaturated fatty acids from diet has been associated with a decreased risk of AD and MCI, and better cognitive aging (12). In addition, high levels of n-3 polyunsaturated fatty acids biomarkers are related to better cognitive functions and higher brain volumes (12).
In conclusion, these four systematic reviews indicates that high-level of consumption of mono- and poly- unsaturated fatty acids and a low consumption of saturated fatty acids could reduce the risk of cognitive decline.

3) We recommend to increase fruit and vegetable intake: GRADE 1B

Fruits and vegetables contain a significant amount of vitamins and minerals, dietary fiber and beneficial non-nutrient substances such as plant sterols, flavonoids and other antioxidants. A large and varied consumption of fruits and vegetables makes it possible to cover the adequate intake of many of these essential nutrients.
This recommendation is based on three systematic reviews, including one meta-analysis, published between 2011 and 2019 (8, 16, 17). First, the meta-analysis of Mottaghi (2018), including six studies showed that increased fruit and vegetable intake was associated with reduced risk of CI (OR: 0.79; 95% CI: 0.67-0.93; P = 0.006) (16). Second, Solfrizzi et al. (2011) concluded that fruit and vegetable consumption is a protective factor for cognitive decline, dementia and AD, despite limited evidence. (8). Third, in the Science project, the authors highlighted that subjects included in the “high veggy” dietary pattern (based on eight nutritional items assessed with a food frequency questionnaire) had a higher MMSE scores (β 0.30 (0.21–0.64)). They concluded that better adherence to a “high veggy” dietary pattern was related to better global cognition (17).
In conclusion, despite the limited epidemiological evidence available, the literature seems to support the positive effect of fruit and vegetable high intake to prevent or treat cognitive impairment.

4) We recommend a greater consumption of vitamin D: GRADE 1C

Vitamin D plays an important role in the regulation of calcium absorption and homeostasis. The major source of vitamin D comes from the natural sun exposure. In addition, the main sources of dietary vitamin D are fish/fish products followed by eggs, fats/oils, bread/bakery products, and milk/dairy product.
The fourth recommendation is based on four recent systematic reviews published between 2015 and 2018 (18-21). Among these papers, three meta-analysis were performed. First, a meta-analysis highlighted severe vitamin D deficiency increased the risk of dementia (<25 nmol/L or 7-28 nmol/L) compared to people with sufficient vitamin D supply (≥50 nmol/L or 54-159 nmol/L) (point estimate 1.54; 95% CI 1.19-1.99, I2 = 20%) (19). Second, in the meta-analysis published by Goodwill and Szoeke (2017), low vitamin D was associated with worse cognitive performance (OR = 1.24, CI = 1.14-1.35) and cognitive decline (OR = 1.26, CI = 1.09-1.23); with a more marked effect in cross-sectional studies than in longitudinal studies. However, no significant difference between vitamin D supplementation and control on cognition was shown (SMD = 0.21, CI = -0.05 to 0.46) (20). Meta-analyses from Krause et al. (2015) have shown that low levels of vitamin D were associated with diminished cognitive function although the available evidence is not sufficient to establish causality (18). However, according to Butler et al. (2018), evidence about effects vitamin D plus calcium supplements was either insufficient, suggesting that these supplements did not reduce risk for cognitive decline (21).
In conclusion, the current literature suggest that low vitamin D levels might contribute to the development of dementia. Observational evidence demonstrates low vitamin D is related to poorer cognition. However, interventional studies still need to show a clear benefit from vitamin D supplementation.

5) We recommend that people who are interested can try a ketogenic diet, but that evidence does not yet support widespread uptake: GRADE 2C

The ketogenic diet is a low-carbohydrate (usually <50 g/day) and fat-rich diet that causes the body to break down fat into molecules called ketones. Ketones circulate in the blood and become the main source of energy for many cells in the body.
This recommendation is based on two recent systematic reviews (22, 23). The results of the review performed by Włodarek (2019) indicate that studies with human participants have demonstrated a reduction of AD symptoms after application of a ketogenic diet (22). These results corroborate those of White et al. (2017) (23). In this systematic review including four studies assessing the role of ketone supplementation in patients with either mild cognitive impairment (MCI) or Alzheimer’s disease, the authors state that ketosis may be beneficial for subjects with AD (23). Finally, Lilamand et al. (2020) have recently conducted a systematic review assessing the effects of Ketogenic diet on cognition (24). Eleven human studies were included in this review and most of these studies showed that ketone supplementation or Ketogenic diet led to an improvement of cognitive outcomes (such as global cognition, memory and executive functions) (24).
In conclusion, the results suggesting the positive short-term effects of Ketogenic diet seems promising. However, neither data on the long-term application of the ketogenic diet in patients with neurodegenerative disease or data on its effects on disease symptoms are available.

6) We suggest a lower consumption of full-fat dairy and/or dairy fats, but that evidence does not yet support widespread uptake: GRADE 2C

In addition to milk, dairy products include yogurt, cheese, etc. This food category is rich in calcium, protein, potassium and phosphorus.
This recommendation is based on two systematic reviews published in 2011 and 2018 (8, 25) and is in line with the Mediterranean diet. In the systematic of Solfrizzi et al., only one out of the five principal prospective reports found a beneficial effect from dairy consumption in regard to cognitive function or MCI/dementia risk. Poorer cognitive function and an increased risk of vascular dementia were found to be associated with a lower consumption of milk or dairy products. On the other hand, three longitudinal studies found adverse effects on cognitive function from dairy fat. Indeed, consumption of whole-fat dairy products may be associated with cognitive decline in the elderly (8). In addition, Lee et al. performed a meta-analysis of three cohort studies and showed no significant association between milk intake and cognitive decline outcome (pooled adjusted risk ratio = 1.21; 95% CI: 0.81, 1.82; the highest vs. the lowest intake) with large statistical heterogeneity (I 2 = 64.1%) (25).
In conclusion, the overall strength of evidence is inadequate for the positive effects of milk or dairy consumption on cognitive decline and disorders while whole-fat dairy products may be associated with late-life cognitive decline.

7) We suggest a caloric restriction, but that evidence does not yet support widespread uptake: GRADE 2C

Caloric restriction could be defined as a reduction in energy intake below the usual calorie consumption (≥10% in human studies and usually 20% or higher in rodent species).
The association between weight loss and cognitive function in older adults is still unclear. Siervo et al. (2011) estimated the effectiveness of intentional weight loss on cognitive function in overweight and obese adults (n=12: seven RCTs & five studies included a control group). The authors concluded that the effect of weight loss on memory (effect size 0.13, 95% CI 0.00-0.26, P=0.04) and on attention/executive functioning (effect size 0.14, 95% CI 0.01-0.27, P<0.001) is small, but significant. Weight loss appears to be associated with low-order improvements in executive/attention functioning and memory in obese, but not in overweight individuals (26).
The review of Gillette-Guyonnet, S. and B. Vellas (2008) corroborates this systematic review, suggesting that caloric restriction could be used to promote successful brain aging (27). In addition, according to Levenson and Rich (2007), caloric restriction has wide-ranging health benefits and may offer protection against age-related neuronal loss and neurodegenerative disorders such as Alzheimer’s disease, possibly via enhanced adult neurogenesis (28).
More recently, Cremonini et al. (2019), stated that short periods of caloric restrictions are able to improve cognitive function (verbal memory) in elderly subjects but, according to these researchers, it is hard to believe that severe restrictions could be tolerated for long periods, especially in elderly subjects affected by neurodegenerative diseases (29).
In conclusion, despite the gaps in our current knowledge, the information we have seems to support the importance of caloric intake on the development and prevention of neurodegenerative disorders (28).

Insufficient evidence for three other nutrients are developed below

In addition to the previous recommendations, more studies are required in other components of dietary patterns in relation to cognition.


They are composed of one or more polypeptide chains. Each of these chains consists of amino acid residues linked together by peptide bonds. Protein can be represented as primary, secondary, tertiary, and quaternary structures. The primary sequence (amino acid) is the sequence of interest from a nutritional point of view.
Dietary protein and its amino acids influence cognition through its effects on risk factors related with cognitive decline. To date, according to the recent review of Glenn et al. (2019), there are limited scientific data directly linking protein/amino acid intake to cognitive decline (30). In this review, Glenn concluded that protein can help maximize physical activity results and physical activity is beneficial for maintaining cognitive status. However, a direct mechanism of action between protein and cognitive status remains unclear. Therefore, further investigations are necessary to understand the underlying mechanism explicating the effect of protein and its constituent amino acids on cognitive health. In 2015, Koh et al. had already performed a systematic review to obtain and appraise relevant studies on the effects of dietary protein or thiamine on cognitive function in healthy older adults (31). Seventeen eligible studies were included in this review and the authors concluded that evidence supporting an association between higher protein and/or thiamine intakes and better cognitive function was weak (31). Moreover, there was no evidence to support the role of specific protein food sources, such as types of meat, on cognitive function (31).
To date, the literature does not allow us to formulate recommendation on protein intake to prevent or manage cognitive decline.

Vitamins B6, B12 and folates

B vitamins are a water-soluble vitamin that play an important role in cell metabolism.
In 2019, Ford and Almeida performed a meta-analysis to examine the efficacy of treatment with vitamin B12, vitamin B6, or folic acid (B9) in slowing cognitive decline among older adults with and without cognitive impairment (32). This meta-analysis of randomized controlled trials including patients with pre-existing cognitive impairment (CI; n=10) or without cognitive impairment (NCI; n=21) found that B-vitamin supplementation does not improve cognitive status (Mini-Mental State Examination scores; CI studies: mean difference 0.16 (95% confidence interval -0.18 to 0.51) vs. NCI studies: mean difference 0.04 (95% confidence interval -0.10 to 0.18)) compared to placebo (32). The results of Ford and Almeida are consistent with those of Forbes et al. published in 2015 (33). In the latest meta-analysis, including seven trials investigating the effects of various combinations of folate, B6 and/or B12 vitamins on cognitive performance, the authors showed no effect of vitamin B on MMSE score (three trials, mean difference 0.02, 95% CI 0.22 to 0.25) (33). In addition, the meta-analysis of Cao et al. (2015), conducted to determine whether there is an association between diet and risk of dementia, highlighted that vitamin B intake was related with decrease of dementia (RR: 0.72, 95 % CI: [0.54-0.96], P = 0.026) (14).
Because of the limitations of the available data, the lack of evidence of effect should not necessarily be interpreted as evidence of no effect. However, the current evidence does not allow us to establish a recommendation for vitamins B6, B9 (or folates) and B12 in the context of cognitive impairments.


An antioxidant is a molecule that prevents or slows down oxidation by neutralizing free radicals. Excess free radicals are responsible for cell damage. Therefore, antioxidants delay or inhibit cellular damage.
Rafnsson et al. (2013) investigated, through a systematic review, the possible cognitive benefits of antioxidant nutrients in the elderly (34). According to the authors, the main supportive evidence came from two high quality studies. The first reported an accelerated decline in global cognition, attention, and psychomotor speed but also a decrease in plasma selenium levels over a 9 years period. The second study observed that people in the higher quartile of intake of vitamin C, E, and carotenes have less cognitive decline during the 3 years of follow-up. All associations persisted after adjustment for confounding factors. Although the evidence is limited in number, the authors concluded that antioxidants have potential protective effect against cognitive decline in older adults. Additional quality investigations are needed to confirm this possible association (34). In the meta-analysis of Forbes et al. (2015), three trials investigated the effect of vitamin E supplementation on cognitive impairment (33). Contrary to the conclusions of Rafsson et al. (34), no statistically significant effect on any of the cognitive outcomes examined was found (33). The conclusions of Crichton et al. (2013) also support those of Forbes. This review, assessing the association between antioxidant intakes (vitamins C, E, flavonoids, carotenoids) and cognitive function or risk for dementia included eight cross-sectional and thirteen longitudinal studies (35). The authors found mixed associations between antioxidant intake, cognition and risk of dementia and Alzheimer’s disease (35).
These inconsistent results do not allow us to establish clear recommendations for antioxidants to prevent or manage cognitive impairments. Future studies are warranted to elucidate the effects of a high intake of dietary antioxidants on cognitive function.



Overall, some nutritional factors appear to either increase the risk of cognitive decline, or protect against it. Risk could be conferred by diets high in milk and dairy products. Of the dietary patterns that appear to offer some protection, the best evidence supports adherence to a Mediterranean diet to decrease the risk of cognitive decline. This evidence also contributes to support of a diet rich in mono- and poly- unsaturated fatty acids, fruit and vegetable, vitamin D and low in saturated fatty acids. Indeed, there is an overlap between the global recommendation related to dietary pattern (Mediterranean diet) and those related to specific nutrients (mono- and poly- unsaturated fatty acids, fruit and vegetable, vitamin D and low in saturated fatty acid). The current literature does not show enough to support more than a ketogenic diet as a promising, but unproved means of managing or preventing cognitive decline. Indeed, the results suggesting the positive short-term effects of Ketogenic diet seems promising. However, there is no evidence for long-term effects of the ketogenic diet among patients with neurodegenerative disease or for its effects on disease symptoms. In the same line, we suggest a lower consumption of whole-fat dairy products, while the existing evidence (mostly observational) is too poor to draw a firm conclusion about the beneficial effect of milk or dairy intake on the risk of cognitive impairments in adults. The literature also seems to support the importance of caloric intake on the development and prevention of neurodegenerative disorders, but that evidence does not yet support widespread uptake. Finally, more studies are required in other components of dietary patterns in relation to cognition, such as protein, vitamins B (B6, B12 and folates) or antioxidants.
Beside these recommendations, eating and drinking difficulties are recognized sources of ill health in people with dementia. A systematic review on importance of nutritional status in people with cognitive impairment or dementia showed: 1) a small positive short-term but unclear long term effects in controlled interventional studies; 2) a small evidence in food modification or dysphagia management studies and; 3) inconsistent evidence in eating assistance studies, whatever the settings, the level of care and support or the type and the degree of dementia (36). In addition, “finger foods” (foods that can be easily eaten with fingers) increase the pleasure of eating, the food consumption and the autonomy among frail older adults or those with dementia (37, 38). Pouyet et al. reported also that finger foods were frequently chosen and consumed by nursing home residents (39). It is also been concluded that finger food is an easy and cheap strategy (40).
In conclusion, it is clear that diet is an important modifiable factor to prevent or protect against cognitive decline. However, more studies are required to determine the recommended duration and amounts of nutrients.


Funding: M. Aubertin-Leheudre and F. Buckinx are supported by the Fonds de Recherche du Québec Santé (FRQS). F. Buckinx is also supported by the IRSC (Instituts de recherche en santé du Canada).

Conflict of interest: The authors have no conflict of interest to declare



1. Peracino A, Pecorelli S. The Epidemiology of Cognitive Impairment in the Aging Population: Implications for Hearing Loss. Audiology & neuro-otology 2016;21 Suppl 1: 3.
2. Ferri CP, Prince M, Brayne C, Brodaty H, Fratiglioni L, Ganguli M, Hall K, Hasegawa K, Hendrie H, Huang Y, Jorm A, Mathers C, Menezes PR, Rimmer E, Scazufca M. Global prevalence of dementia: a Delphi consensus study. Lancet (London, England) 2005;366: 2112.
3. McGrattan AM, McEvoy CT, McGuinness B, McKinley MC, Woodside JV. Effect of dietary interventions in mild cognitive impairment: a systematic review. The British journal of nutrition 2018;120: 1388.
4. Rochoy M, Rivas V, Chazard E, Decarpentry E, Saudemont G, Hazard PA, Puisieux F, Gautier S, Bordet R. Factors Associated with Alzheimer’s Disease: An Overview of Reviews. The journal of prevention of Alzheimer’s disease 2019;6: 121.
5. WHO. Risk reduction of cognitive decline and dementia. In, France. 2019.
6. Ogawa S. Nutritional management of older adults with cognitive decline and dementia. Geriatrics & gerontology international 2014;14 Suppl 2: 17.4
7. Yusufov M, Weyandt LL, Piryatinsky I. Alzheimer’s disease and diet: a systematic review. The International journal of neuroscience 2017;127: 161.
8. Solfrizzi V, Panza F, Frisardi V, Seripa D, Logroscino G, Imbimbo BP, Pilotto A. Diet and Alzheimer’s disease risk factors or prevention: the current evidence. Expert review of neurotherapeutics 2011;11: 677.
9. Samadi M, Moradi S, Moradinazar M, Mostafai R, Pasdar Y. Dietary pattern in relation to the risk of Alzheimer’s disease: a systematic review. Neurological sciences : official journal of the Italian Neurological Society and of the Italian Society of Clinical Neurophysiology, 2019
10. Brozek JL, Akl EA, Alonso-Coello P, Lang D, Jaeschke R, Williams JW, Phillips B, Lelgemann M, Lethaby A, Bousquet J, Guyatt GH, Schunemann HJ. Grading quality of evidence and strength of recommendations in clinical practice guidelines. Part 1 of 3. An overview of the GRADE approach and grading quality of evidence about interventions. Allergy 2009;64: 669.
11. Aurora RN, Zak RS, Auerbach SH, Casey KR, Chowdhuri S, Karippot A, Maganti RK, Ramar K, Kristo DA, Bista SR, Lamm CI, Morgenthaler TI. Best practice guide for the treatment of nightmare disorder in adults. Journal of clinical sleep medicine : JCSM : official publication of the American Academy of Sleep Medicine 2010; 6: 389.
12. Otaegui-Arrazola A, Amiano P, Elbusto A, Urdaneta E, Martinez-Lage P. Diet, cognition, and Alzheimer’s disease: food for thought. European journal of nutrition 2014;53: 1.
13. Aridi YS, Walker JL, Wright ORL. The Association between the Mediterranean Dietary Pattern and Cognitive Health: A Systematic Review. Nutrients 2017;9.
14. Cao GY, Li M, Han L, Tayie F, Yao SS, Huang Z, Ai P, Liu YZ, Hu YH, Xu B. Dietary Fat Intake and Cognitive Function among Older Populations: A Systematic Review and Meta-Analysis. The journal of prevention of Alzheimer’s disease2019; 6: 204.
15. Zhang Y, Chen J, Qiu J, Li Y, Wang J, Jiao J. Intakes of fish and polyunsaturated fatty acids and mild-to-severe cognitive impairment risks: a dose-response meta-analysis of 21 cohort studies. The American journal of clinical nutrition 2016;103: 330.
16. Mottaghi T, Amirabdollahian F, Haghighatdoost F. Fruit and vegetable intake and cognitive impairment: a systematic review and meta-analysis of observational studies. European journal of clinical nutrition 2018;72: 1336.
17. Wesselman LMP, Doorduijn AS, de Leeuw FA, Verfaillie SCJ, van Leeuwenstijn-Koopman M, Slot RER, Kester MI, Prins ND, van de Rest O, de van der Schueren MAE, Scheltens P, Sikkes SAM, van der Flier WM. Dietary Patterns Are Related to Clinical Characteristics in Memory Clinic Patients with Subjective Cognitive Decline: The SCIENCe Project. Nutrients 2019;11.
18. Krause D, Roupas P. Effect of Vitamin Intake on Cognitive Decline in Older Adults: Evaluation of the Evidence. The journal of nutrition, health & aging 2015;19: 745.
19. Sommer I, Griebler U, Kien C, Auer S, Klerings I, Hammer R, Holzer P, Gartlehner G. Vitamin D deficiency as a risk factor for dementia: a systematic review and meta-analysis. BMC geriatrics 2017;17: 16.
20. Goodwill AM, Szoeke C. A Systematic Review and Meta-Analysis of The Effect of Low Vitamin D on Cognition. Journal of the American Geriatrics Society2017; 65: 2161.
21. Butler M, Nelson VA, Davila H, Ratner E, Fink HA, Hemmy LS, McCarten JR, Barclay TR, Brasure M, Kane RL. Over-the-Counter Supplement Interventions to Prevent Cognitive Decline, Mild Cognitive Impairment, and Clinical Alzheimer-Type Dementia: A Systematic Review. Annals of internal medicine 2018;168: 52.
22. Wlodarek D. Role of Ketogenic Diets in Neurodegenerative Diseases (Alzheimer’s Disease and Parkinson’s Disease). Nutrients 2019;11.
23. White; H, Venkatesh; K, Venkatesh B. Systematic Review of the Use of Ketones in the Management of Acute and Chronic Neurological Disorders. Journal of neurology and neuroscience 2017;8: 188.
24. Lilamand M, Porte B, Cognat E, Hugon J, Mouton-Liger F, Paquet C. Are ketogenic diets promising for Alzheimer’s disease? A translational review. Alzheimer’s research & therapy 2020;12: 42.
25. Lee J, Fu Z, Chung M, Jang DJ, Lee HJ. Role of milk and dairy intake in cognitive function in older adults: a systematic review and meta-analysis. Nutrition journal 2018;17: 82.
26. Siervo M, Arnold R, Wells JC, Tagliabue A, Colantuoni A, Albanese E, Brayne C, Stephan BC. Intentional weight loss in overweight and obese individuals and cognitive function: a systematic review and meta-analysis. Obesity reviews : an official journal of the International Association for the Study of Obesity 2011;12: 968,.
27. Gillette-Guyonnet S, Vellas B. Caloric restriction and brain function. Current opinion in clinical nutrition and metabolic care 2008;11: 686.
28. Levenson CW, Rich NJ. Eat less, live longer? New insights into the role of caloric restriction in the brain. Nutrition reviews 2007;65: 412.
29. Cremonini AL, Caffa I, Cea M, Nencioni A, Odetti P, Monacelli F. Nutrients in the Prevention of Alzheimer’s Disease. Oxidative medicine and cellular longevity 2019: 9874159.
30. Glenn JM, Madero EN, Bott NT. Dietary Protein and Amino Acid Intake: Links to the Maintenance of Cognitive Health. Nutrients 2019;11.
31. Koh F, Charlton K, Walton K, McMahon AT. Role of dietary protein and thiamine intakes on cognitive function in healthy older people: a systematic review. Nutrients 2015;7: 2415.
32. Ford AH, Almeida OP. Effect of Vitamin B Supplementation on Cognitive Function in the Elderly: A Systematic Review and Meta-Analysis. Drugs & aging 2019;36: 419.
33. Forbes SC, Holroyd-Leduc JM, Poulin MJ, Hogan DB. Effect of Nutrients, Dietary Supplements and Vitamins on Cognition: a Systematic Review and Meta-Analysis of Randomized Controlled Trials. Canadian geriatrics journal : CGJ 2015;18: 231.
34. Rafnsson SB, Dilis V, Trichopoulou A. Antioxidant nutrients and age-related cognitive decline: a systematic review of population-based cohort studies. European journal of nutrition 2013;52: 1553.
35. Crichton GE, Bryan J, Murphy KJ. Dietary antioxidants, cognitive function and dementia–a systematic review. Plant foods for human nutrition (Dordrecht, Netherlands) 2013;68: 279,.
36. Abdelhamid A, Bunn D, Copley M, Cowap V, Dickinson A, Gray L, Howe A, Killett A, Lee J, Li F, Poland F, Potter J, Richardson K, Smithard D, Fox C, Hooper L. Effectiveness of interventions to directly support food and drink intake in people with dementia: systematic review and meta-analysis. BMC geriatrics 2016;16: 26.
37. Roberts HC, Lim SER, Cox NJ, Ibrahim K. The Challenge of Managing Undernutrition in Older People with Frailty. Nutrients 2019;11.
38. Heelan M, Prieto J, Roberts H, Gallant N, Barnes C, Green S. The use of finger foods in care settings: an integrative review. Journal of human nutrition and dietetics : the official journal of the British Dietetic Association 2020;33: 187.
39. G.Cuvelier VPAGLB. Attractiveness and consumption of finger foods in elderly Alzheimer’s disease patients. Food Quality and Preference 2014;34: 62.
40. Cluskey M, Kim YK. Use and perceived effectiveness of strategies for enhancing food and nutrient intakes among elderly persons in long-term care. Journal of the American Dietetic Association 2001;101: 111,.



S.A. Galle1,2,*, I.K. Geraedts1,*, J.B. Deijen1,3, M.V. Milders1, M.L. Drent1,4


1. Department of Clinical, Neuro- & Developmental Psychology, Clinical Neuropsychology Section, Vrije Universiteit Amsterdam, Amsterdam, The Netherlands; 2. Department of Epidemiology, Erasmus Medical Centre, Rotterdam, the Netherlands; 3. Hersencentrum Mental Health Institute, Amsterdam, The Netherlands; 4. Department of Internal Medicine, Section of Endocrinology, Amsterdam UMC, location VUmc, Amsterdam, The Netherlands * These authors contributed equally to this work

Corresponding Author: Sara A. Galle, Department of Clinical, Neuro- & Developmental Psychology, Clinical Neuropsychology Section, Vrije Universiteit Amsterdam, Van der Boechorststraat 7-9, 1081 BT Amsterdam, The Netherlands, T: 0031205988769, E-mail:

J Prev Alz Dis 2020;4(7):265-273
Published online March 2, 2020,



Aging is associated with a decrease in body and brain function and with a decline in insulin-like growth factor 1 levels. The observed associations between alterations in insulin-like growth factor 1 levels and cognitive functioning and Mild Cognitive Impairment suggest that altered insulin-like growth factor 1 signaling may accompany Alzheimer’s disease or is involved in the pathogenesis of the disease. Recent animal research has suggested a possible association between insulin-like growth factor 1 levels and the Apolipoprotein E ε4 allele, a genetic predisposition to Alzheimer’s disease. It is therefore hypothesized that a reduction in insulin-like growth factor 1 signaling may moderate the vulnerability to Alzheimer’s disease of human Apolipoprotein E ε4 carriers. We address the impact of age-related decline of insulin-like growth factor 1 levels on physical and brain function in healthy aging and Alzheimer’s disease and discuss the links between insulin-like growth factor 1 and the Apolipoprotein E ε4 polymorphism. Furthermore, we discuss lifestyle interventions that may increase insulin-like growth factor 1 serum levels, including physical activity and adherence to a protein rich diet and the possible benefits to the physical fitness and cognitive functioning of the aging population.

Key words: Insulin-like growth factor, Alzheimer’s disease, ApoE-ε4 allele, physical activity, diet, aging.



It is well known that the process of aging is associated with physical and mental changes. In the body, normal aging is primarily associated with a decrease in muscle mass and strength. In the brain, normal aging is mainly characterized by metabolic changes in the prefrontal cortex and associated with a decrease in brain size and synaptic plasticity (1). These changes in body and brain lead to alterations in physical, as well as cognitive functioning in elderly people, such as increased frailty and decreased cognitive performance (1, 2).
When age-related cognitive decline becomes qualitatively severe and progresses rapidly, it is likely to progress into a clinical diagnosis of dementia. The most common form of dementia is Alzheimer’s disease (AD). While there are some medications that decelerate the neuropathological progression of AD or offer some symptomatic relief, there is no cure available. In the absence of a cure for AD, research has focused on the most common risk factors and preventive strategies. Important non-modifiable risk factors for AD that have been investigated include age and genetics. Potentially modifiable factors are risk factors that are associated with lifestyle like socioeconomic factors, diet, cerebrovascular disease, and physical inactivity (3).
In the development of preventive strategies, it is important to understand the interplay between
neurobiological and lifestyle factors. One important factor that is both influenced by lifestyle factors like physical activity and diet (4, 5) and plays a role in the maintenance of physical fitness (6) and cognitive functioning (7) is insulin-like growth factor 1 (IGF-1). This review will discuss the impact of age-related decline of IGF-1 levels on physical and cognitive functioning in healthy aging and AD. In addition, we discuss the possible link between IGF-1 and ApoE-ε4. Furthermore, we explore how lifestyle interventions focusing on physical activity and diet may be useful to improve physical fitness and cognitive functioning by increasing IGF-1 serum levels.
Insulin-like growth factor 1 is a peptide growth hormone, with a structure similar to insulin, encoded by the IGF-1 gene located on chromosome 12. As part of the growth hormone (GH)/ IGF-1 axis, IGF-1 plays an essential role in growth of the body and development of the brain. IGF-1 is mainly produced in the liver, stimulated by GH, which is secreted from the anterior pituitary gland. IGF-1 can also be produced in local peripheral tissues such as muscle and bone tissue when GH binds to its Growth Hormone Receptor (GHR) (8). As IGF-1 is GH dependent and, unlike GH, circulating IGF-1 levels do not fluctuate widely over time, IGF-1 is a more reliable measure and appropriate marker for GH status (9). Therefore, this review focuses on neurobiological processes and lifestyle factors related to IGF-1.


IGF-1 and the aging body

Throughout the body, IGF-1 regulates the development and function of cells. It promotes cell growth and contributes to cell proliferation, stress resistance and survival in many cell types (10). IGF-1 can bind with high affinity to the IGF-1 receptor (IGF-1R), but also to the insulin receptor (11) as its structure is closely related to insulin. The IGF-1R is expressed in many distinct tissues in the body. For this reason IGF-1 can have different effects, such as the promotion of neuronal survival in the central nervous system and the facilitation of peripheral muscle regeneration (12). Because of the essential role of IGF-1 in muscle growth and the involvement of IGF-1 in many mechanisms and functions of the body, IGF-1 is an important factor for embryonic and childhood growth (13) and anabolic processes in adults (14).
Aging is associated with a decline in IGF-1 (10). The progressive decline has been termed the ‘somatopause’, which may be caused by potential alterations of the hypothalamic regulation of GH secretion, in particular an age-dependent decrease in endogenous hypothalamic GHRH output, contributing to the age-associated GH and IGF-1 decline (15). Moreover, low physical fitness and higher adiposity in older individuals also contribute to the decreased GH secretion and associated IGF-1 decline (16). Low levels of IGF-1 are associated with decreased skeletal muscle mass and function (17). Studies have shown that IGF-1 serum levels are positively associated with muscular strength and walking speed and are negatively associated with self-reported difficulty in mobility tasks (18). Systemic infusion of GH over 8 hours led to increased GH and IGF-1 concentration levels and increased muscle protein synthesis in eight healthy young adults aged 18 to 24 years (19). In addition, Rudman et al. (20) demonstrated increased lean body mass, average vertebral bone density, IGF-1 levels, and decreased body fat following GH administration over 6 months in nine healthy adults that were not observed in 12 untreated healthy adult men. Mauras et al. (14]) used recombinant human IGF-1 (rhIGF-1) treatment to increase IGF-1 plasma levels in 10 patients with Laron’s syndrome, characterized by GH receptor deficiency, and showed that increased IGF-1 plasma levels were associated with increased lean body mass and decreased fat mass. Furthermore, Dik et al. (21) demonstrated that higher IGF-1 serum levels were associated with fewer functional limitations (e.g. difficulties with climbing stairs, cutting toenails, use of public transport) in 1318 healthy participants aged 65 to 88 years. This association suggests that reduced IGF-1 levels in older people might make them more prone to these functional limitations.
The influence of IGF-1 on bone development has been demonstrated using mouse models. Bikle et al. (22) found a 24% decrease in cortical bone size and reduced femoral lengths, but increased connectivity and trabecular bone density, in IGF 1 deficient (Igf-1 -/-) mice. In addition, a study by Courtland et al. (23) used inducible liver IGF-1 deficient mice to deplete IGF-1 serum levels at varying times in mice development and demonstrated that depletion of serum IGF-1 levels at four weeks in male mice resulted in reduced trabecular and cortical bone acquisition by 16 weeks. Depletion of serum IGF-1 levels in mice of eight weeks resulted in decreased cortical bone properties at 32 weeks, whereas depletion of IGF-1 serum levels after peak bone acquisition at 16 weeks did not lead to detrimental effects on bone.
Finkenstedt et al. (24) demonstrated that 12 months of recombinant human GH (rhGH) treatment of 18 adult male and female patients, with adult onset GH deficiency, and an average age of 44 years, resulted in increased markers of bone formation and resorption and elevated IGF-1 levels compared to the untreated group. Following rhGH treatment for 12 months, markers for bone turnover, including bone formation and resorption, increased relative to baseline in those patients who were treated with rhGH. In addition, after 12 months, IGF-1 was significantly increased in all patients treated with rhGH, and bone mineral density in the lumbar and proximal spine was increased in this group, particularly in patients with low bone mass. Furthermore, one month of recombinant human GH administration in 10 healthy older men, with an average age of 68 years, led to improved balance and stair climb time as well as increased muscle IGF-1 gene expression (25). Ohlsson et al. (26) also showed that low IGF-1 serum levels in elderly men were associated with increased risk of bone fractures (e.g. hip, spine), which are partly caused by falls and are a clear marker of physical frailty. Muscle weakness, functional limitations, and age are substantial contributors to the risk of falls in elderly and these factors are all associated with a decrease in IGF-1. Hence, the age-related decrease in IGF-1 may play an important role in the increased incidence of falls in elderly.


IGF-1 and the aging brain

IGF-1 produced by the liver has the ability to cross the blood-brain barrier and can subsequently bind to IGF-1 receptors expressed throughout the brain. High densities of IGF-1 receptors are observed in various brain areas including the amygdala, thalamic nuclei, hippocampus, superficial and deep cortical layers, olfactory bulb, cerebellum, cerebral cortex, caudate nucleus, frontal cortex and the putamen (27). In addition, IGF-1 is also produced in brain tissues and can thereby act locally via paracrine or autocrine mechanisms. IGF-1 plays an important role in neuronal growth, the maintenance of synapses and the protection of neurons in the brain (28). Furthermore, IGF-1 has been found to enhance and maintain myelination, essential for the propagation of neuronal impulses, in the central nervous system (CNS) as well as in the peripheral nervous system.
Age-related decline of IGF-1 levels is associated with altered brain function. Sonntag et al. (29) showed age-related decreases in IGF-1 receptor density in hippocampal and cortical regions in rats. The authors found that IGF-1 mRNA levels were reduced in the cerebellum in older rats, compared to younger ones. This decline was associated with an increase in cell death (30). As IGF-1 is involved in maintaining myelination in the CNS, age-related IGF-1 decline may be associated with the breakdown of myelination which in turn may have a negative impact on cognition in humans (31). This age-related breakdown of myelin can lead to decreased signal transmission speed in neurons, essential for integration of information between highly distributed neural networks that underlie higher cognitive functions, such as executive processing (32).


IGF-1, cognition and MCI

Evidence thus far has supported the idea that IGF-1 plays an essential role in cognition. In healthy men and women, IGF-1 serum levels have been shown to be positively related to working memory (33), selective attention, executive function (34), verbal fluency and performance on the Mini-Mental State Examination (MMSE) (35). A recent study by Maass et al. (36) demonstrated that an increase in IGF-1 serum levels was positively associated with hippocampal volume and verbal memory recall in a population of healthy elderly. In childhood-onset GH deficient men GH substitution improved both mood and memory. These improvements were maintained during the 10 year follow-up period (37).
With respect to pathological cognitive aging, IGF-1 levels have been found to be reduced in people with MCI compared to cognitively healthy people. MCI is associated with reduced performance in various cognitive domains, including attention, executive function, processing speed, visuospatial skill and memory. Doi et al. (38) conducted a population survey in 3355 participants with an average age of 71.4 years and found that people with MCI showed decreased IGF-1 serum levels compared to cognitively healthy people. Furthermore, Calvo et al. (39) showed a positive association between IGF-1 serum levels and cognitive performance, mainly in the domains of learning and memory, in elderly people with MCI, suggesting IGF-1 may be neuroprotective in elderly people susceptible to AD. This notion is supported by the finding that the cognitive impairments in AD may be partly related to reduced IGF-1 serum levels (40).


IGF-1 and AD

At a neurobiological level AD is characterized by several neurotoxic effects caused by senile plaques (SPs) and neurofibrillary tangles (NFTs) that lead to synaptic dysfunction, neuronal cell death and cerebral atrophy, mainly in the hippocampus and temporal and parietal lobes. The main elements of SPs are beta-amyloid (Aβ) aggregates. These Aβ aggregates form plaques outside neurons that intervene with communication between neurons at synapses and contribute to neuronal cell death. NFTs, on the other hand, are primarily composed of hyperphosphorylated tau protein. Deviant abnormal tau proteins inside neurons (tau tangles) block the transports of essential molecules, such as nutrients in the neuron, thereby contributing to cell death. The abundance of NFTs is positively associated with the severity of AD (41). These brain alterations impede the transfer of information between synapses and cause a reduction in the number of synapses. The progression of the disease eventually leads to neuronal cell death causing a substantial shrinkage of the brain.
In 2007, Alvarez et al. (40) showed subnormal IGF-1 levels in adults diagnosed with AD. Additionally, Westwood et al. (42) showed that lower IGF-1 serum levels are associated with an increased risk of developing AD in older- and middle-aged people. This study also demonstrated that higher levels of IGF-1 are associated with greater brain volumes, even among cognitively healthy older and middle-aged people, suggesting a protective effect of IGF-1 against neurodegeneration. Recent evidence showed that IGF-1 resistance in the brain is increased in AD (43). Moloney et al. (44) demonstrated that alterations in IGF-1 receptors (IGF-1Rs) in the AD temporal cortex, including reduced expression as well as an aberrant distribution of IGF-1Rs in the neurons, contribute to impaired IGF-1R signaling in AD neurons. The deviant distribution of IGF-1Rs in neurons away from the plasma membrane suggests that IGF-1Rs are less able to respond to extracellular IGF-1 in AD, contributing to possible IGF-1R signaling resistance in neurons that degenerate (44). A decrease in IGF-1 signaling can contribute to loss of myelin function, which is thought to result in nerve fiber conduction delays found in people with AD (45). Furthermore, deficits in IGF-1 signaling have been related directly to AD pathology like increased accumulation of Aβ, phosphorylated tau, increased neuro-inflammation and apoptosis (28), suggesting that impaired IGF-1 signaling plays a role in the pathogenesis of AD. In contrast to this idea it has also been suggested that downregulation of IGF-1 signaling is a consequence of neuropathology and alterations in IGF-1 signaling could be seen as a compensatory response to attenuate the effects of aging and neurodegeneration. This idea is supported by the assocation between suppression of IGF-1 signaling and longevity in humans (46) and the observation that low IGF-1 levels predict life expectancy in exceptionally long-lived individuals (47).
In model organisms in which IGF-1 signaling was attenuated increased lifespan and a delayed process of aging has been observed (48, 49). For instance, in AD mouse models the long-term suppression of IGF-1 signaling reduced neuronal loss and increased resistance to oxidative stress and neuro-inflammation. In line with these findings, lowerd IGF-1 serum levels in transgenic mouce models, induced by a protein restriction diet, alleviated AD pathology (50).
In human observational studies, a recent meta-analysis by Ostrowski and colleagues could not confirm the hypothesized association between serum IGF-1 and AD. From 3540 studies that analyzed the relation between IGF-1 and AD, only 10 studies provided serum IGF-1 values. These 10 studies included 850 AD patients and 871 controls. From these studies 5 reported that AD subjects had higher IGF-1 levels, 2 reported no difference in IGF-1 levels and 3 reported lower IGF-1 levels in AD. The authors conclude that serum IGF-1 may be a personalized factor reflecting differential influence of genetic polymorphisms, age of onset or disease progression of AD patients (51). It is important to note that the number of included studies poses limitations to the generalizability of the results and more studies are needed to clarify the possible relationship between IGF-1 levels and AD.


Potential interactions of IGF-1 and ApoE-ε4 in the development of AD

The Apolipoprotein E gene, APOE, is the largest genetic risk factor associated with cognitive decline in late-onset AD (52). ApoE is involved in lipid transport in the central and peripheral nervous system, and brain injury repair. The three most common alleles of APOE (ε2, ε3, ε4) encode for the three major isoforms (ApoE-ε2, ApoE-ε3, ApoE-ε4) of the apolipoprotein E (ApoE), a protein that plays a central role in brain injury repair, lipid transport and metabolism. The ε2, ε3 and ε4 alleles have a worldwide frequency of 8.4%, 77.9% and 13.7%, respectively (53).
The strength of the effects of the different APOE genotypes on AD risk differs between ethnic groups. In the present study, we will focus on Caucasians. ApoE-ε3 is often considered the neutral allele with regard to AD risk. Compared to the ApoE-ε3, ApoE-ε4 is associated with both an increased incidence rate and an earlier onset of AD. One copy of ApoE-ε4 increases the risk of developing AD threefold, while those who are homozygous for ε4 have an approximately 13-fold increased risk (54). ApoE-ε4 carriers also have an enhanced risk for developing vascular dementia and mild cognitive impairment (MCI) (55) and studies have shown that the ApoE-ε4 allele is involved in the acceleration of cognitive decline (56). The accelerated cognitive decline observed in ApoE-ε4 carriers could be an important clinical precursor of AD. It has been shown that ApoE promotes the proteolytic breakdown of the Aβ aggregates appearing in AD, whereas the isoform ApoE-ε4 is less effective in enhancing this breakdown (57). Moreover, Kumar et al. (58) demonstrated that neurofibrillary tangle density was increased in ApoE-ε4 carriers relative to non-carriers of the allele. Hence, carrying the ApoE-ε4 allele increases the vulnerability of the brain to AD pathology.
As described earlier, IGF-1 has an opposite effect to ApoE-ε4 on N-methyl-D-aspartate receptor (NMDAR) signaling and Aβ clearance in the brain (59). With respect to NMDAR signaling, Liu et al. (60) demonstrated that the ApoE-ε4 allele enhanced an age-related decline in cognitive function in mice by decreasing NR2B subunit levels which in turn down-regulates the NMDAR pathway. Specifically, NR2B may play a role in spatial learning and long-term potentiation (61, 62). In contrast, IGF-1 has been found to positively affect the NMDARr pathway in rats by increasing NR2B subunits (62).
Impairments in Aβ clearance are a major hallmark in early as well as late AD. People carrying the ApoE-ε4 allele are more vulnerable to disturbances in Aβ clearance than people not carrying this allele (63). IGF-1 supports Aβ clearance in the healthy brain (64).
A recent study by Keeney et al. (65) was the first to report a direct association between the three isoforms of ApoE (ε2, ε3 and ε4) and IGF-1 by demonstrating deficient IGF-1 gene expression and reduced IGF-1 protein level in mice carrying the human ApoE-ε3 and ApoE-ε4, compared to mice carrying the human ApoE-ε2 allele. This association indicates that the three isoforms of ApoE affect IGF-1 signaling differently, suggesting a potential mechanism that might contribute to the differences in AD risk of ApoE isoforms (65).
Moderation of the association between IGF-1 signaling and AD by APOE genotype has previously been suggested in experimental studies. Using microarray analysis of the astrocyte transcriptome, Simpson and colleagues demonstrated that as AD pathology progresses, downregulation of gene transcription in astrocytes leads to a reduction in the expression of intra-cellular insulin and IGF signaling pathways, particularly in individuals expressing the ApoE-ε4 allele (66). Impaired IGF-1 signaling in human astrocytes is associated with a reduced ability to protect neurons from oxidative stress, which has been identified as an important factor in the promotion of tau and Aβ pathology in AD (67).
Therapeutic approaches targeting insulin resistance by increasing IGF-1, insulin, or insulin sensitivity have been promising, but do suggest differential effects in people with or without genetic susceptibility to AD. More specifically, intravenous and intranasal insulin administration in patients with AD, reduced amyloid precursor protein (APP) levels and improved memory scores only in those without the ApoE-ε4 allele (68, 69).
Previously, our group reported tentative evidence of an interaction between the ApoE-ε4 allele and IGF-1 receptor stimulating activity in an elderly cohort (59). IGF-1 receptor stimulating activity in the median and top tertiles was related with increased dementia incidence in hetero- and homozygotes of the ApoE-ε4 allele, but did not show any association with dementia risk in people without the ApoE-ε4 allele (59). The observed elevation in IGF-1 receptor stimulating activity may have marked a compensatory response to neuropathological changes associated with the ApoE-ε4 genotype. Additionally, we found that the ApoE-ε4 homozygotes, with a lifetime risk of Alzheimer’s Disease of 80% (70), have the lowest IGF-1 levels (59). Similarly, a genome-wide association study on longevity by Deelen et al. (2001) showed that the ApoE-ε4 allele was related to lower IGF-1 levels in middle-aged women. Hence, the increased risk of developing AD in ApoE-ε4 carriers might partially be attributed to alterations in IGF-1 signaling (71).


Physical activity and IGF-1

As mentioned earlier, IGF-1 serum levels can be influenced by lifestyle factors, such as physical activity (5). Aerobic and anaerobic exercise interventions have been shown to influence IGF-1 levels. The positive effect of aerobic exercise on IGF-1 levels has been shown in a mouse study that demonstrated upregulated mRNA levels of IGF-1 in mice after 15 days of voluntary wheel running. Protein levels of IGF-1 in the dentate gyrus had also increased (72). Replication of these results in human participants was provided by several studies that showed an increase in IGF-1 serum levels following aerobic exercise in adults (73, 74). Likewise, a study concerning the effect of anaerobic exercise on IGF-1 serum levels reported positive effects of anaerobic training on IGF-1 levels in healthy older men (75). There is, however, still much controversy concerning the association between physical exercise and IGF-1 levels. A systematic review of experimental studies on the effect of physical activity on measures of IGF-1 and cognitive functioning in healthy elderly concluded moderate intensity aerobic training and moderate and high intensity resistance training may improve circulating IGF-1 and cognition, depending on the sex of the participant and duration of the training. However, disparities in the type of exercise, protocols and samples hinder comparison of the results and the establishment of consensus (76).
Furthermore, negative associations between IGF-1 levels and physical activity, could also be explained by favorable neuromuscular anabolic adaption, which is a normal short-term adaptive response of the body to increased physical exercise (Rarick et al., 2007). It has been thought that during episodes of active muscle building IGF-1 serum levels decrease (78), but local muscle gene expression and production of IGF-1 increase (79). Longitudinal studies on exercise interventions indicate that IGF-1 serum levels may only decline temporarily and may increase after longer duration of intensive training and are maintained when training is reduced (74). The long-term effect of physical activity on IGF-1 levels may be explained by epigenetic alterations. It is known that physical activity can contribute to changes in various physiological systems by epigenetic mechanisms (80). Physical activity may induce epigenetic modifications to the IGF-1 gene, leading to sustained increased IGF-1 levels (6, 80). There is evidence showing that these types of alterations can be inherited (81). In light of epigenetics and the influence of prolonged physical activity on IGF-1 levels, the current decrease in the number of physically active people, mainly in high-income countries, is alarming.
Regular engagement in physical activity could be of special importance to those with a genetic susceptibility to AD. Several studies have indicated that the negative association between regular physical activity and cognitive decline is limited to those with one or more copies of the ApoE-ε4 allele. Schuit et al. registered engagement in physical activity in a group of elderly Dutch men and found that while risk of cognitive decline did not differ between active and inactive ApoE-ε4 non-carriers the risk was 4 times higher in inactive ApoE-ε4 carriers compared to active ApoE-ε4 carriers (82). A similar finding, indicating that inactivity is especially detrimental to cognitive abilities for ApoE-ε4 carriers, was reported in a longitudinal study in a Finnish cohort (83). Rovio et al. found a significant relationship between physical activity at midlife and risk of developing AD at a 21-year follow-up for ApoE-ε4 carriers, but not for ApoE-ε4 non-carriers. Additionally, Kivipelto et al. (84) demonstrated that physical inactivity increased the risk of AD mainly among ApoE-ε4 carriers.
Several brain-imaging studies have reported support for these findings. Deeny et al. found that in the middle-aged, sedentary ApoE-ε4 carriers exhibited lower activity levels in the temporal lobe, a region known to be vulnerable to early decline in AD, relative to active ApoE-ε4 carriers, while activity level did not distinguish between AD risk for ApoE-ε4 non carriers (85). In 2012 Head et al. demonstrated that in cognitively normal older adults those who were sedentary and ApoE-ε4 carriers showed more Aβ deposition than active ApoE-ε4 carriers, whereas this association was not present in non-carriers (86). Subsequently, Smith et al. observed that the hippocampal volume of those ApoE-ε4 carriers that displayed low levels of physical activity was on average 3% lower in comparison to non-carriers, and in comparison to ApoE-ε4 carriers who displayed high levels of physical activity (87), indicating that physical inactivity may be related to brain atrophy in ApoE-ε4 carriers. Together, these studies suggest that ApoE-ε4 carriers may be more susceptible to the negative effects of physical inactivity, and that sedentary ApoE-ε4 carriers may be at increased risk of developing AD.
In contrast, in a functional MRI study Smith et al. observed that among ApoE-ε4 carriers being engaged in higher levels of physical activity was associated with greater regional brain activation during a semantic memory task in comparison to non-carriers and ApoE-ε4 carriers who displayed lower levels of physical activity (88), suggesting that ApoE-ε4 carriers do not suffer more from inactivity than any other group but do experience more benefits from physical activity.
On the other hand, studies have shown that the interaction between physical activity and cognitive decline is restricted to ApoE-ε4 non-carriers. In a prospective study among older adults Podewils et al. found an inverse association between physical activity and risk of AD after a 5 year follow-up that was confined to ApoE-ε4 non-carriers, indicating that benefits of exercise may be confined only to ε4 non-carriers (89). A similar finding was reported after a 5 year follow-up in cognitively healthy elders (90). Fenesi et al. found a significant protective effect of physical activity regarding dementia risk in ApoE-ε4 non-carriers, and no significant effect in ApoE-ε4 carriers. One randomly controlled trial supported these two observational studies (91). Lautenschlager et al. studied the effect of an exercise intervention on cognitive functioning in a randomized trial in healthy older adults with subjective memory impairment. The researchers found a modest improvement in cognitive functioning in those treated with the intervention. In a post-hoc comparison, treatment response interacted with APOE genotype, as ApoE-ε4 non-carriers showed a significantly larger improvement compared to both carriers and non-carriers in the control condition, while no other significant differences were found (91).
One study did not find a significant interaction effect between physical activity and cognition and ApoE-ε4 carrier status (92). Luck et al. failed to find an interaction between physical activity in late life and risk of AD in an observational study after a 4.5-year follow-up in a group of healthy elderly aged 75 years and over. However, the authors did note that the interaction between ApoE-ε4 and low physical activity for AD risk verged on the border of significance.
With regard to physical fitness, it has been found that the presence of the ApoE-ε4 allele is associated with motor decline (e.g. motor performance) in older people (93) and the strength of this relationship increases with age. Further analysis showed that this association was mainly due to a greater age-related decrease in upper and lower limb muscle strength in people carrying the ApoE-ε4 allele. This study showed that ApoE-ε4 carriers are at greater risk of rapid motor decline relative to non-carriers, particularly later in life. Considering that limited physical activity is associated with motor decline, and physical activity is potentially protective against cognitive decline, physical activity is argued to be especially relevant to ApoE-ε4 carriers (86, 93).


Diet and IGF-1

In addition to the effect of physical activity on IGF-1 levels, diet is an important lifestyle factor affecting IGF-1 levels. Norat et al. (4) demonstrated that protein intake was positively associated with IGF-1 serum levels. This study showed that intake of milk, calcium, magnesium, phosphorus, potassium, vitamin B6, and vitamin B2 was positively related to IGF-1 serum levels and that the intake of vegetables and beta-carotene was negatively associated with IGF-1 serum levels in women. In line with this study, a study by Allen et al. (94) demonstrated that in adult women aged 20 to 70 a plant-based (vegan) diet was related to lower IGF-1 serum levels compared to women with a meat-eating or lacto-ovo-vegetarian diet. The difference in IGF-1 serum levels between the groups was mainly explained by protein intake consisting of essential amino acids. Long-term caloric restriction for a duration of 1 and 6 years was not associated with with reduced IGF-1 serum levels in healthy middle aged men and women, if protein intake is high (95). In addition, a recent study by Fontana et al. (96) showed that 2 years of caloric restriction did not affect IGF-1 serum levels in healthy non-obese young and middle-aged men and women, suggesting no sustained effects of caloric restriction on IGF-1 serum levels. Though, other studies demonstrated that short term caloric restriction for 6 days lowers IGF-1 serum levels (97), indicating that particularly short term fasting lowers IGF-1 serum levels.


Exercise combined with diet and IGF-1

Few studies have examined the influence of physical activity combined with a specific diet on IGF-1 levels. A negative caloric balance induced by physical exercise or caloric restriction, were both associated with equivalent decline in IGF-1 levels (98). Smith et al. (98) concluded that a decline in IGF-1 levels is mainly explained by an energy deficit, irrespective whether this deficit was induced by caloric restriction or physical exercise. A study by Rarick et al. (77) demonstrated a decline in IGF-1 serum levels after 7 days of increased physical activity in healthy men. However, the decrease in IGF-1 serum level was not moderated by fitness intensity, energy balance, or dietary protein intake. This study therefore challenges the concept of Smith et al. (98)and suggests that yet unknown mechanisms related to physical activity, such as enhanced energy flux, may affect IGF-1 levels independently.


IGF-1 in relation to other AD risk factors

When investigating the association between IGF-1 and Alzheimer’s disease it is important to consider the limited role of epidemiological evidence in causal inference and the possible confounding influence of a myriad of factors that are related to both AD risk and altered IGF-1 signaling. Among these potential confounders are lifestyle factors, like alcohol and nicotine consumption (99–101), and several conditions associated with alterations in insulin or IGF-1 signaling such as type 2 diabetes, obesity, cardiovascular disease, cerebral infarcts (102–107) and depression (108, 109). These cross-links between altered IGF-1 signaling and increased risk of AD highlight the importance of experimental and meta-analytic evidence, replication studies and a thorough consideration of potential confounders in the association between IGF-1 signaling and Alzheimer’s disease.


Conclusion and future perspectives

Although there are contradictory findings on the association between physical exercise, diet and IGF-1 it can be argued that promoting physical activity and a protein rich diet could be promising interventions that may increase IGF-1 levels, thereby increasing physical fitness and counteracting age-related neurodegeneration and AD. Further research, including experimental, epidemiological and multi-omic approaches (110), is warranted to investigate the prospective value of different biomarker profiles for future dementia risk. Findings can be applied to improve early diagnostics and to increase the efficiency of lifestyle interventions targeting IGF-1 signaling to delay or prevent the development of physical and cognitive decline, in particular for those most vulnerable for AD.



– IGF-1 is associated with cognitive deficits and pathological alterations in the brain that accompany AD
– Decreased IGF-1 levels are a possible moderator of genetic vulnerability to AD
– Increasing physical activity and adherence to a protein rich diet may be useful interventions to increase IGF- serum levels, thereby increasing physical fitness and cognitive functioning


Funding: The authors received no financial support for the research, authorship or publication of this manuscript.

Conflict of interests: All authors declare that they have no conflict of interest.

Open Access: This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits use, duplication, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.



1. Peters R. Ageing and the brain. Postgrad Med J 2006;82:84–8.
2. Clegg A, Young J, Iliffe S, et al. Frailty in elderly people. Lancet 2013;381:752–62.
3. Keene CD, Montine TJ. Epidemiology, pathology, and pathogenesis of Alzheimer disease. In: Post DW (ed) Up to date. UpToDate, Waltham, MA, 2015;pp 1–18
4. Norat T, Dossus L, Rinaldi S, et al. Diet, serum insulin-like growth factor-I and IGF-binding protein-3 in European women. Eur J Clin Nutr 2007;61:91–8.
5. Moran S, Chen Y, Ruthie A, Nir Y. Alterations in IGF-I affect elderly: Role of physical activity. Eur Rev Aging Phys Act 2007;4:1–8.
6. Eliakim A, Brasel JA, Mohan S, et al. Physical fitness, endurance training, and the growth hormone-insulin-like growth factor I system in adolescent females. J Clin Endocrinol Metab 1996;81:3986–92.
7. Arwert LI, Deijen JB, Drent ML. The relation between insulin-like growth factor I levels and cognition in healthy elderly: a meta-analysis. Growth Horm IGF Res 2005;15:416–22.
8. Salvatori R. Growth hormone and IGF-1. Rev Endocr Metab Disord 2004;5:15–23.
9. McMahon CD, Chai R, Radley-Crabb HG, et al. Lifelong exercise and locally produced insulin-like growth factor-1 (IGF-1) have a modest influence on reducing age-related muscle wasting in mice. Scand J Med Sci Sports 2014;24:e423-435.
10. Ashpole NM, Sanders JE, Hodges EL, et al. Growth hormone, insulin-like growth factor-1 and the aging brain. Exp Gerontol 2015;68:76–81.
11. Lesniak MA, Hill JM, Kiess W, et al. Receptors for insulin-like growth factors I and II: autoradiographic localization in rat brain and comparison to receptors for insulin. Endocrinology 1988;123:2089–99.
12. Duan C, Ren H, Gao S. Insulin-like growth factors (IGFs), IGF receptors, and IGF-binding proteins: Roles in skeletal muscle growth and differentiation. Gen Comp Endocrinol 2010;167:344–351.
13. Kajantie E, Fall CHD, Seppälä M, et al. Serum insulin-like growth factor (IGF)-I and IGF-binding protein-1 in elderly people: relationships with cardiovascular risk factors, body composition, size at birth, and childhood growth. J Clin Endocrinol Metab 2003;88:1059–65.
14. Mauras N, Martinez V, Rini A, Guevara-Aguirre J. Recombinant human insulin-like growth factor I has significant anabolic effects in adults with growth hormone receptor deficiency: studies on protein, glucose, and lipid metabolism. J Clin Endocrinol Metab 2000;85:3036–42.
15. Russell-Aulet M, Jaffe CA, Demott-Friberg R, Barkan AL. In vivo semiquantification of hypothalamic growth hormone-releasing hormone (GHRH) output in humans: Evidence for relative GHRH deficiency in aging. J Clin Endocrinol Metab 1999;84:3490–3497.
16. Vahl N, Jørgensen JOL, Jurik AG, Christiansen JS. Abdominal adiposity and physical fitness are major determinants of the age associated decline in stimulated GH secretion in healthy adults. J Clin Endocrinol Metab 1996;81:2209–2215.
17. Clemmons DR. Role of IGF-I in skeletal muscle mass maintenance. Trends Endocrinol Metab 2009;20:349–356.
18. Cappola AR, Bandeen-Roche K, Wand GS, et al. Association of IGF-I levels with muscle strength and mobility in older women. J Clin Endocrinol Metab 2001;86:4139–46.
19. Fryburg DA, Barrett EJ, Louard R, al. et. Growth hormone acutely stimulates skeletal muscle but not whole-body protein synthesis in humans. Metabolism 1993;42:1223–7.
20. Rudman D, Feller AG, Nagraj HS, et al. Effects of Human Growth Hormone in Men over 60 Years Old. N Engl J Med 1990;323:1–6.
21. Dik MG, Pluijm SM., Jonker C, et al. Insulin-like growth factor I (IGF-I) and cognitive decline in older persons. Neurobiol Aging 2003;24:573–581.
22. Bikle D, Majumdar S, Laib A, et al . The skeletal structure of insulin-like growth factor I-deficient mice. J Bone Miner Res 2001;16:2320–9.
23. Courtland HW, Elis S, Wu Y, et al. Serum IGF-1 affects skeletal acquisition in a temporal and compartment-specific manner. PLoS One 2011;6:1–10.
24. Finkenstedt G, Gasser RW, Höfle G, et al. Effects of growth hormone (GH) replacement on bone metabolism and mineral density in adult onset of GH deficiency: results of a double-blind placebo-controlled study with open follow-up. Eur J Endocrinol 1997;136:282–9
25. Brill KT, Weltman AL, Gentili A, et al. Single and Combined Effects of Growth Hormone and Testosterone Administration on Measures of Body Composition, Physical Performance, Mood, Sexual Function, Bone Turnover, and Muscle Gene Expression in Healthy Older Men. J Clin Endocrinol Metab 2002;87:5649–5657.
26. Ohlsson C, Mellström D, Carlzon D, et al. Older men with low serum IGF-1 have an increased risk of incident fractures: the MrOS Sweden study. J Bone Miner Res 2011;26:865–72.
27. Adem A, Jossan SS, d’Argy R, et al. Insulin-like growth factor 1 (IGF-1) receptors in the human brain: quantitative autoradiographic localization. Brain Res 1989;503:299–303
28. Bedse G, Di Domenico F, Serviddio G, Cassano T. Aberrant insulin signaling in Alzheimer’s disease: Current knowledge. Front Neurosci 2015;9:.
29. Sonntag WE, Lynch CD, Bennett SA, et al. Alterations in insulin-like growth factor-1 gene and protein expression and type 1 insulin-like growth factor receptors in the brains of ageing rats. Neuroscience 1999;88:269–79
30. Pañeda C, Arroba AI, Frago LM, et al. Growth hormone-releasing peptide-6 inhibits cerebellar cell death in aged rats. Neuroreport 2003;14:1633–5.
31. O’Sullivan M, Jones DK, Summers PE, et al. Evidence for cortical “disconnection” as a mechanism of age-related cognitive decline. Neurology 2001;57:632–638.
32. Bartzokis G. Age-related myelin breakdown: a developmental model of cognitive decline and Alzheimer’s disease. Neurobiol Aging 2004;25:5–18.
33. Deijen JB, Arwert LI, Drent ML. The GH/IGF-I Axis and Cognitive Changes across a 4-Year Period in Healthy Adults. ISRN Endocrinol 2011;1–6.
34. Bellar D, Glickman EL, Juvancic-Heltzel J, Gunstad J. Serum insulin like growth factor-1 is associated with working memory, executive function and selective attention in a sample of healthy, fit older adults. Neuroscience 2011;178:133–137.
35. Al-Delaimy WK, von Muhlen D, Barrett-Connor E. Insulinlike growth factor-1, insulinlike growth factor binding protein-1, and cognitive function in older men and women. J Am Geriatr Soc 2009;57:1441–6.
36. Maass A, Düzel S, Brigadski T, et al. Relationships of peripheral IGF-1, VEGF and BDNF levels to exercise-related changes in memory, hippocampal perfusion and volumes in older adults. Neuroimage, 2015.
37. Arwert LI, Deijen JB, Müller M, Drent ML. Long-term growth hormone treatment preserves GH-induced memory and mood improvements: A 10-year follow-up study in GH-deficient adult men. Horm Behav 2005;47:343–349.
38. Doi T, Shimada H, Makizako H, et al. Association of insulin-like growth factor-1 with mild cognitive impairment and slow gait speed. Neurobiol Aging 2015;36:942–7.
39. Calvo D, Gunstad J, Miller LA, et al. Higher serum insulin-like growth factor-1 is associated with better cognitive performance in persons with mild cognitive impairment. Psychogeriatrics 2013;13:170–4.
40. Alvarez A, Cacabelos R, Sanpedro C, et al. Serum TNF-alpha levels are increased and correlate negatively with free IGF-I in Alzheimer disease. Neurobiol Aging 2007;28:533–6.
41. Hardy J, Selkoe DJ. The amyloid hypothesis of Alzheimer’s disease: progress and problems on the road to therapeutics. Science 2002;297:353–6.
42. Westwood AJ, Beiser A, Decarli C, et al. Insulin-like growth factor-1 and risk of Alzheimer dementia and brain atrophy. Neurology 2014;82:1613–9.
43. Talbot K, Wang H-Y, Kazi H, et al. Demonstrated brain insulin resistance in Alzheimer’s disease patients is associated with IGF-1 resistance, IRS-1 dysregulation, and cognitive decline. J Clin Invest 2012;122:1316–1338.
44. Moloney AM, Griffin RJ, Timmons S, et al. Defects in IGF-1 receptor, insulin receptor and IRS-1/2 in Alzheimer’s disease indicate possible resistance to IGF-1 and insulin signalling. Neurobiol Aging 2010;31:224–43.
45. de la Monte SM. Contributions of brain insulin resistance and deficiency in amyloid-related neurodegeneration in Alzheimer’s disease. Drugs 2012;72:49–66.
46. Suh Y, Atzmon G, Cho M-O, et al. Functionally significant insulin-like growth factor I receptor mutations in centenarians. Proc Natl Acad Sci U S A 2008;105:3438–42.
47. Milman S, Atzmon G, Huffman DM, et al. Low insulin-like growth factor-1 level predicts survival in humans with exceptional longevity. Aging Cell 2014;13:769–771.
48. Tazearslan C, Huang J, Barzilai N, Suh Y (2011. Impaired IGF1R signaling in cells expressing longevity-associated human IGF1R alleles. Aging Cell 2011;10:551–554.
49. Tatar M, Kopelman A, Epstein D, et al. A mutant Drosophila insulin receptor homolog that extends life-span and impairs neuroendocrine function. Science 2001(80- ) 292:107–110.
50. Parrella E, Maxim T, Maialetti F, et al. Protein restriction cycles reduce IGF-1 and phosphorylated tau, and improve behavioral performance in an Alzheimer’s disease mouse model. Aging Cell 2013;12:257–268.
51. Ostrowski PP, Barszczyk A, Forstenpointner J, et al. Meta-analysis of serum insulin-like growth factor 1 in Alzheimer’s disease. PLoS One 2016;11:1–12.
52. Rohn TT. Proteolytic cleavage of apolipoprotein E4 as the keystone for the heightened risk associated with Alzheimer’s disease. Int J Mol Sci 2013;14:14908–22.
53. Farrer LA. Effects of Age, Sex, and Ethnicity on the Association Between Apolipoprotein E Genotype and Alzheimer Disease. JAMA 1997;278:1349.
54. Farrer LA, Cupples LA, Haines JL, et al. Effects of age, sex, and ethnicity on the association between apolipoprotein E genotype and Alzheimer disease. A meta-analysis. JAMA 1997;278:1349–56.
55. Liu C-C, Liu C-C, Kanekiyo T, et al. Apolipoprotein E and Alzheimer disease: risk, mechanisms and therapy. Nat Rev Neurol 2013;9:106–18.
56. Marioni RE, Campbell A, Scotland G, et al. Differential effects of the APOE e4 allele on different domains of cognitive ability across the life-course. Eur J Hum Genet 2015;24:919–923.
57. Hori Y, Hashimoto T, Nomoto H, et al. Role of Apolipoprotein E in β-Amyloidogenesis. J Biol Chem 2015;290:15163–15174.
58. Kumar NT, Liestøl K, Løberg EM, et al. Apolipoprotein E allelotype is associated with neuropathological findings in Alzheimer’s disease. Virchows Arch 2015;467:225–35.
59. Galle SA, Van Der Spek A, Drent ML, et al. Revisiting the role of insulin-like growth factor-I receptor stimulating activity and the apolipoprotein E in Alzheimer’s disease. Front Aging Neurosci 2019;10:1–9.
60. Liu D, Pan X, Zhang J, et al. APOE4 enhances age-dependent decline in cognitive function by down-regulating an NMDA receptor pathway in EFAD-Tg mice. Mol Neurodegener 2015;10:7.
61. Chen QS, Kagan BL, Hirakura Y, Xie CW. Impairment of hippocampal long-term potentiation by Alzheimer amyloid beta-peptides. J Neurosci Res 2000;60:65–72
62. Le Grevès M, Le Grevès P, Nyberg F. Age-related effects of IGF-1 on the NMDA-, GH- and IGF-1-receptor mRNA transcripts in the rat hippocampus. Brain Res Bull 2005;65:369–374.
63. Wildsmith KR, Holley M, Savage JC, et al. Evidence for impaired amyloid β clearance in Alzheimer’s disease. Alzheimers Res Ther 2013;5:33.
64. Gasparini L, Xu H. Potential roles of insulin and IGF-1 in Alzheimer’s disease. Trends Neurosci 2003;26:404–406.
65. Keeney JT-R, Ibrahimi S, Zhao L. Human ApoE Isoforms Differentially Modulate Glucose and Amyloid Metabolic Pathways in Female Brain: Evidence of the Mechanism of Neuroprotection by ApoE2 and Implications for Alzheimer’s Disease Prevention and Early Intervention. J Alzheimer’s Dis 2015;48:411–424.
66. Simpson JE, Ince PG, Shaw PJ, et al. Microarray analysis of the astrocyte transcriptome in the aging brain: Relationship to Alzheimer’s pathology and APOE genotype. Neurobiol Aging 2011;32:1795–1807.
67. Huang WJ, Zhang X, Chen WW. Role of oxidative stress in Alzheimer’s disease (review). Biomed Reports 2016;4:519–522.
68. Craft S, Asthana S, Schellenberg G, et al. Insulin effects on glucose metabolism, memory, and plasma amyloid precursor protein in Alzheimer’s disease differ according to apolipoprotein-E genotype. Ann N Y Acad Sci 2000;903:222–228.
69. Reger MA, Watson GS, Frey WH, et al. Effects of intranasal insulin on cognition in memory-impaired older adults: Modulation by APOE genotype. Neurobiol Aging 2006;27:451–458.
70. van der Lee SJ, Wolters FJ, Ikram MK, et al. The effect of APOE and other common genetic variants on the onset of Alzheimer’s disease and dementia: a community-based cohort study. Lancet Neurol 2018;17:434–444.
71. Deelen J, Beekman M, Uh HW, et al. Genome-wide association study identifies a single major locus contributing to survival into old age; the APOE locus revisited. Aging Cell 2011;10:686–698.
72. Yu J-L, Ma L, Ma L, Tao Y-Z. [Voluntary wheel running enhances cell proliferation and expression levels of BDNF, IGF1 and WNT4 in dentate gyrus of adult mice]. Sheng Li Xue Bao 2014;66:559–68
73. Ardawi M-SM, Rouzi AA, Qari MH. Physical activity in relation to serum sclerostin, insulin-like growth factor-1, and bone turnover markers in healthy premenopausal women: a cross-sectional and a longitudinal study. J Clin Endocrinol Metab 2012;97:3691–9.
74. Koziris LP, Hickson RC, Chatterton RT, et al. Serum levels of total and free IGF-I and IGFBP-3 are increased and maintained in long-term training. J Appl Physiol 1999;86:1436–42
75. Amir R, Ben-Sira D, Sagiv M. Igf-I and fgf-2 responses to wingate anaerobic test in older men. J Sports Sci Med 2007;6:227–32
76. Stein AM, Silva TMV, Coelho FG de M, et al. Physical exercise, IGF-1 and cognition A systematic review of experimental studies in the elderly. Dement Neuropsychol 2018;12:114–122.
77. Rarick KR, Pikosky MA, Grediagin A, et al. Energy flux, more so than energy balance, protein intake, or fitness level, influences insulin-like growth factor-I system responses during 7 days of increased physical activity. J Appl Physiol 2007;103:1613–21.
78. Nindl BC, Headley SA, Tuckow AP, et al. IGF-I system responses during 12 weeks of resistance training in end-stage renal disease patients. Growth Horm IGF Res 220414:245–50.
79. Schmitz KH, Ahmed RL, Yee D. Effects of a 9-month strength training intervention on insulin, insulin-like growth factor (IGF)-I, IGF-binding protein (IGFBP)-1, and IGFBP-3 in 30-50-year-old women. Cancer Epidemiol Biomarkers Prev 2002;11:1597–604
80. Ntanasis-Stathopoulos J, Tzanninis JG, Philippou A, Koutsilieris M. Epigenetic regulation on gene expression induced by physical exercise. J Musculoskelet Neuronal Interact 2013;13:133–46
81. Jablonka E, Lamb MJ. The inheritance of acquired epigenetic variations. Int J Epidemiol 2915;44:1094–103.
82. Schuit AJ, Feskens EJM, Launer LJ, Kromhout D. Physical activity and cognitive decline, the role of the apolipoprotein e4 allele. Med Sci Sports Exerc 2001;33:772–777
83. Rovio S, Kåreholt I, Helkala E-L, et al. Leisure-time physical activity at midlife and the risk of dementia and Alzheimer’s disease. Lancet Neurol 2005;4:705–711.
84. Kivipelto M, Rovio S, Ngandu T, et al. Apolipoprotein E epsilon4 magnifies lifestyle risks for dementia: a population-based study. J Cell Mol Med 2008;12:2762–71.
85. Deeny SP, Poeppel D, Zimmerman JB, et al. Exercise, APOE, and working memory: MEG and behavioral evidence for benefit of exercise in epsilon4 carriers. Biol Psychol 2008;78:179–187.
86. Head D, Bugg JM, Goate AM, et al. Exercise engagement as a moderator of the effects of APOE genotype on amyloid deposition. Arch Neurol 2012;69:636.
87. Smith JC, Nielson K a., Woodard JL, et al. Physical activity reduces hippocampal atrophy in elders at genetic risk for Alzheimer’s disease. Front Aging Neurosci 2014;6:1–7.
88. Smith JC, Nielson KA, Woodard JL, et al. Interactive effects of physical activity and APOE-e4 on BOLD semantic memory activation in healthy elders. Neuroimage 2011;54:635–644.
89. Podewils LJ, Guallar E, Kuller LH, et al. Physical activity, APOE genotype, and dementia risk: Findings from the Cardiovascular Health Cognition Study. Am J Epidemiol 2005;161:639–651.
90. Fenesi B, Fang H, Kovacevic A, et al. Physical exercise moderates the relationship of apolipoprotein E (APOE) genotype and dementia risk: a population-based study. J Alzheimer’s Dis 2017;56:297–303.
91. Lautenschlager NT, Cox KL, Flicker L, et al. Effect of physical activity on cognitive function in older adults at risk for Alzheimer disease: A randomized trial. JAMA – J Am Med Assoc 2008;300:1027–1037.
92. Luck T, Riedel-Heller SG, Luppa M, et al. Apolipoprotein E epsilon 4 genotype and a physically active lifestyle in late life: analysis of gene-environment interaction for the risk of dementia and Alzheimer’s disease dementia. Psychol Med 2014;44:1319–29.
93. Buchman AS, Boyle PA, Wilson RS, et al. Apolipoprotein E e4 allele is associated with more rapid motor decline in older persons. Alzheimer Dis Assoc Disord 2009;23:63–9
94. Allen NE, Appleby PN, Davey GK, et al. The associations of diet with serum insulin-like growth factor I and its main binding proteins in 292 women meat-eaters, vegetarians, and vegans. Cancer Epidemiol Biomarkers Prev 2002;11:1441–8
95. Fontana L, Weiss EP, Villareal DT, et al. Long-term effects of calorie or protein restriction on serum IGF-1 and IGFBP-3 concentration in humans. Aging Cell 2008;7:681–7
96. Fontana L, Villareal DT, Das SK, et al. Effects of 2-year calorie restriction on circulating levels of IGF-1, IGF-binding proteins and cortisol in nonobese men and women: a randomized clinical trial. Aging Cell 2016;15:22–7.
97. Smith WJ, Underwood LE, Clemmons DR. Effects of caloric or protein restriction on insulin-like growth factor-I (IGF-I) and IGF-binding proteins in children and adults. J Clin Endocrinol Metab 1995;80:443–9.
98. Smith AT, Clemmons DR, Underwood LE, et al. The effect of exercise on plasma somatomedin-C/insulinlike growth factor I concentrations. Metabolism 1987;36:533–537.
99. Durazzo TC, Mattsson N, Weiner MW. Smoking and increased Alzheimer’s disease risk: A review of potential mechanisms, 2014.
100. Tong M, Yu R, Deochand C, De La Monte SM. Differential contributions of alcohol and the nicotine-derived nitrosamine ketone (NNK) to insulin and insulin-like growth factor resistance in the adolescent rat brain. Alcohol Alcohol 2014;50:670–679.
101. De La Monte SM, Tong M, Cohen AC, et al. Insulin and insulin-like growth factor resistance in alcoholic neurodegeneration. Alcohol Clin Exp Res 2008;32:1630–1644.
102. Li X, Song D, Leng SX. Link between type 2 diabetes and Alzheimer’s disease: From epidemiology to mechanism and treatment. Clin Interv Aging 2015;10:549–560.
103. Talbot K, Wang H, Kazi H, et al. Demonstrated brain insulin resistance in Alzheimer’s disease patients is associated with IGF-1 resistance, IRS-1 dysregulation, and cognitive decline. J Clin Invest 2012;122:1316–1338.
104. Steen E, Terry BM, Rivera EJ, et al. Impaired insulin and insulin-like growth factor expression and signaling mechanisms in Alzheimer’s disease–is this type 3 diabetes? J Alzheimers Dis 2005;7:63–80.
105. Vagelatos NT, Eslick GD. Type 2 diabetes as a risk factor for Alzheimer’s disease: The confounders, interactions, and neuropathology associated with this relationship. Epidemiol Rev 2013;35:152–160.
106. Spielman LJ, Little JP, Klegeris A. Inflammation and insulin/IGF-1 resistance as the possible link between obesity and neurodegeneration. J. Neuroimmunol. 2014;273:8–21
107. Rochoy M, Rivas V, Chazard E, et al. Factors Associated with Alzheimer’s Disease: An Overview of Reviews. J Prev Alzheimer’s Dis 2019;6:121–134.
108. Paslakis G, Blum WF, Deuschle M. Intranasal insulin-like growth factor I (IGF-I) as a plausible future treatment of depression. Med Hypotheses 2012;79:222–225.
109. Mueller PL, Pritchett CE, Wiechman TN, et al. Antidepressant-like effects of insulin and IGF-1 are mediated by IGF-1 receptors in the brain. Brain Res Bull 2018;143:27–35.
110. Bateman RJ, Blennow K, Doody R, et al. Plasma Biomarkers of AD Emerging as Essential Tools for Drug Development: An EU/US CTAD Task Force Report. J Prev Alzheimer’s Dis 2019;6:169–173.

What do you want to do ?

New mail

What do you want to do ?

New mail

What do you want to do ?

New mail

What do you want to do ?

New mail



F. Sierra


Correspondance author: Division of Aging Biology, National Institute on Aging (NIA), National Institutes of Health (NIH) – 7201 Wisconsin Ave, Suite 3N300, Bethesda MD 20892, USA,

J Prev Alz Dis 2020;(7) in press
Published online December 5, 2019,

Key words: Geroscience, aging, Alzheimer’s, neurodegeneration.



With the relentless aging of the population worldwide, two major concerns need our immediate attention: the expected dramatic increase in disease and disability burden, and the decrease in the ratio of working individuals vs. retirees. A comprehensive approach involving experts in many disciplines will be required to tackle these issues. Here we will concentrate on the role of basic biological science in averting the increase in disease and disability, using Alzheimer’s Disease (AD) as a model.
The spectrum of AD represents a serious threat and a psychological burden on people at all ages. The US Alzheimer’s Association estimates that 5.8 million Americans are currently living with Alzheimer’s, and 1 out of 3 seniors dies with the disease or other dementias. Because of its insidious effect on both the individual and his/her surroundings, as well as the associated healthcare cost, AD has been singled out for special efforts by funding agencies and scientists alike, and while some progress has been made, it is clearly not sufficient. Indeed, in the past few decades, scientists have been able to identify the molecular composition of the telltale plaques and tangles and have created a large number of mouse models that, to some extent, recapitulate the pathological characteristics of the disease (1). Furthermore, early efforts identified some major drivers of the familial (rare) form of the disease, and more recently, a large number of genes suspected to play a role on the late-onset, non-familial form of the disease have also been identified (2). The role of these new genes is just beginning to be unraveled. Yet a cure has been elusive, and even prevention strategies have been less than hoped for.
One aspect of the etiology of the disease that has been largely neglected until recent years is the role of aging. It is not by mere chance that so many chronic diseases appear simultaneously, in many species, as individuals reach approximately 2/3 of the lifespan for their species (around 60 years for humans). Most of these chronic diseases differ dramatically from the diseases we were able to conquer in the 20th century, in that they are not caused by external agents such as pathogens and poor environmental quality but rather, they are the result of failures within our own organism. For that reason, these diseases have proven to be less tractable, and fighting them is more complex. But the age-dependency in the appearance of symptoms from multiple chronic diseases belies the fact that aging is by far the major risk factor for most of these chronic conditions (3), including Alzheimer’s disease (4). Importantly, the fact that such diseases occur in multiple species at different chronological times (days in flies, months in mice, years in humans), but always at the same physiological time (about 2/3 of the expected lifespan) indicates that it is the process of aging, not the passage of time, that is central. The passage of time indeed allows the accumulation of damage that can lead to disease and disability. However, this accumulation is often rather slow while the organism is young and resilient. It is only after the process of aging starts weakening that resilience that serious accumulation of damage – and thus disease – occurs. So, it is not simply that as we age, damage has accumulated to an extent that causes disease; rather, it is that as we age, we had lost part of our defenses, thus allowing the damage to accumulate. Taking AD as an example, we know that even individuals with the worst genetic predisposition to the disease won’t develop symptoms when they are toddlers or teenagers, they will develop them late in life (earlier than other, non-genetically afflicted populations, but usually not earlier than their 40s or 50s) (5). Yet, because of their genetic burden, they are producing enormous amounts of deleterious aggregation-prone proteins from before birth! Minimal accumulation and no disease occurs because, while young, their resilience capacity allows them to counteract this burden, and resolve much of the damage through proteostasis mechanisms.
This is not unique to AD, and a similar argument can be brought to bear in many other chronic diseases of the elderly, including cancer, cardiovascular, chronic kidney disease, etc. This is the central tenet of the new field of geroscience: since aging is at the core, and the most important risk factor for so many chronic diseases and conditions, it follows that addressing aging will produce a better outcome than addressing each disease individually (6). Indeed, it is expected that, by slowing down the pace of aging we can delay all such chronic ailments, all at the same time. This is nothing new, since we have always known about the fragility and illnesses that often accompany old age. What is new is the amazing advancements we have had in the last couple of decades in our efforts to understand the biological underpinnings of the aging process. Indeed, scientists have now identified a handful of molecular and cellular pathways that drive the process of aging (7, 8). Moreover, those discoveries have led to the identification of pharmacological and dietary means to slow down aging processes, and some of these interventions are already being tried in the clinic, including rapamycin (9) and senolytics (10).
The AD field has been slow to recognize these developments, but changes are being implemented, among others, through several new initiatives promoted by the US National Institute of Aging, aimed at promoting research into the geroscience underpinnings of AD. In fact, pre-clinical data in various mouse models of AD suggest that interventions aimed at slowing down the aging process might be effective in delaying or slowing down disease progression. As in other diseases that affect preferentially the elderly population, these pre-clinical interventions have focused primarily on rapamycin (11, 12) and senolytics (13). In fact, a strong argument has been made to test rapamycin in clinical trials of the disease (14), and two small phase I trials of senolytics are being planned for the near future (J. Kirkland, pers. comm.). In the accompanying paper by Guerville et al., a vigorous argument is made for the inclusion of geroscience principles in our fight to conquer Alzheimer’s disease. Importantly, the paper also outlines specific areas where attention to the pillars of aging might be fruitful in our efforts against Alzheimer’s.


Disclosures: Dr. Sierra has no conflicts to disclose. The ideas discussed represent Dr. Sierra’s views and do not represent the views of the U.S. government.



1. Hardy J. A hundred years of Alzheimer’s disease research. Neuron. 2006;52:3–13.
2. Naj A.C., Schellenberg G.D. and the Alzheimer’s Disease Genetics Consortium (ADGC). Genomic variants, genes, and pathways of Alzheimer’s disease: An overview. Am J Med Genet B Neuropsychiatr Genet. 2017;174:5-26.
3. Sierra F. and Kohanski R. Geroscience and the trans-NIH Geroscience Interest Group, GSIG. Geroscience 39:1-5.
4. Qiu C., Kivipelto M. and Strauss E. Epidemiology of Alzheimer’s disease: Occurrence, determinants and strategies toward intervention. Dialogues Clin. Neurosci. 2009;11:111-128.
5. Ryman D.C., Acosta-Baena N., Aisen P.S. et al. Symptom onset in autosomal dominant Alzheimer’s disease: a systematic review and meta-analysis. Neurology 2014;83:253-260.
6. Austad SN. The geroscience hypothesis: Is it possible to change the rate of aging? in Advances in Geroscience, Sierra & Kohanski eds., Springer, 2016;pp1-36.
7. López-Otín C., Blasco M.A., Partridge L. et al. The hallmarks of aging. Cell 2013;153:1194-217.
8. Kennedy B.K., Berger S, Brunet A. et al. Geroscience: linking aging to chronic disease. Cell 2014;159:709-713.
9. Mannick JB, Del Giudice G, Lattanzi M et al. mTOR inhibition improves immune function in the elderly. Sci Transl Med. 2014;6:268ra179
10. Justice JN, Nambiar AM, Tchkonia T et al. Senolytics in idiopathic pulmonary fibrosis: Results from a first-in-human, open-label, pilot study. EBioMedicine. 2019;40:554-563.
11. Caccamo A., Majumder S., Richardson A. et al. Molecular interplay between mammalian target of rapamycin (mTOR), amyloid-ß and tau: Effects on cognitive impairments. J. Biol. Chem. 2010;285:13107-13120.
12. Lin A-L., Jahrling J.B., Zhang W. et al. Rapamycin rescues vascular, metabolic and learning deficits in apolipoprotein E4 transgenic mice with presymptomatic Alzheimer’s disease. J. Cereb. Blood Flow Metab. 2017;37:217-226.
13. Bussian T.J., Aziz A., Meyer C.F. et al. Clearance of senescent glial cells prevents tau-dependent pathology and cognitive decline. Nature 2018;562:578-582.
14. Kaeberlein M. and Galvan V. Rapamycin and Alzheimer’s disease: time for a clinical trial? Sci. Transl. Med. 2019;11, eaar4289



Y.H.W. Tsui-Caldwell1, T.J. Farrer7, Z. McDonnell3, Z. Christensen3, C. Finuf3, E.D. Bigler1,2, J.T. Tschanz4,5, M.C. Norton4,5,6, K.A. Welsh-Bohmer7


1. Department of Psychology, Brigham Young University, Provo, UT, USA; 2. Neuroscience Center, Brigham Young University, Provo, UT, USA; 3. Department of Physiology & Development, Brigham Young University, Provo, UT, USA; 4. Department of Psychology, Utah State University, Logan, UT, USA; 5. Center for Epidemiologic Studies, Utah State University, Logan, UT, USA; 6. Department of Family, Consumer & Human Development, Utah State University, Logan, UT, USA; 7. Department of Psychiatry and Neurology, Duke University, Durham, NC

Corresponding Author: Thomas J. Farrer Ph.D., 932 Morreene Road, Durham, NC 27710, USA, (919)-668-6802,

J Prev Alz Dis 2019;(6) in press
Published online January 4, 2019,



BACKGROUND: White matter integrity in aging populations is associated with increased risk of cognitive decline, dementia diagnosis, and mortality. Population-based data can elucidate this association.
Objectives: To examine the association between white matter integrity, as measured by a clinical rating scale of hyperintensities, and mental status in older adults including advanced aging.
Design: Scheltens Ratings Scale was used to qualitatively assess white matter (WM) hyperintensities in participants of the Cache County Memory Study (CCMS), an epidemiological study of Alzheimer’s disease in an exceptionally long-lived population. Further, the relation between Mini-Mental State Exam (MMSE) and WM hyperintensities were explored.
Method: Participants consisted of 415 individuals with dementia and 22 healthy controls.
Results: CCMS participants, including healthy controls, had high levels of WM pathology as measured by Scheltens Ratings Scale score.  While age did not significantly relate to WM pathology, higher Scheltens Ratings Scale scores were associated with lower MMSE findings (correlation between -0.14 & -0.22; p < .05).
Conclusions: WM pathology was common in this county-wide population sample of those ranging in age from 65 to 106. Increased WM burden was found to be significantly associated with decreased overall MMSE performance.

Key words: Scheltens Rating Scale, Cache County, aging, clinical ratings, white matter hyperintensity.



Numerous studies have shown the presence of increased WMHs to be associated with not only cognitive decline but degenerative disorders like AD (16). Given that mental status testing (e.g., MMSE) is often used as a surrogate of general cognitive functioning, multiple studies have examined the influence of WM integrity on MMSE and general neurocognitive and clinical outcomes. Otsuka et al. (17) previously demonstrated that a decline in MMSE scores in older adults is correlated with white matter integrity (diffuse anisotropy and fractional anisotropy) in the diffuse deep hemispheric white matter and in the corpus callosum, even after controlling for white matter lesion volume and total white matter volume. Quinque and colleagues (18) also demonstrated that scores from visual ratings of white matter change are correlated with performance on multiple processing speed tasks. Studies consistently demonstrate that reduced white matter integrity on diffusion tensor parameters is associated with cognitive functioning in elderly samples both with and without cognitive dysfunction (19).  The presence of white matter burden is also associated with increased risk of dementia (20, 21). As such, further exploration of the impact of white matter integrity in older adults with and without cognitive decline into advanced age (75 years+) is warranted to delineate factors associated with the relationship between white matter changes and cognition. Understanding the factors that mediate the relationship between white matter change and cognition has important implications for future interventional approaches in Alzheimer’s disease, both to delay onset of symptoms and to slow progression of disease once manifest.
The Cache County Memory Study (CCMS) dataset has provided important insights related to aging and new onset dementia studies (22). To date, there has not been a comprehensive description of WM findings in the entire Waves 1 and 2 CCMS sample. Likewise, the Scheltens Rating Scale has become one of the standards in the field and the early reports involving the Wave 1 Cohort did not use the Scheltens Rating Scale. Accordingly, this investigation identified WMHs based on the Scheltens Rating Scale in the CCMS population-based sample of individuals with cognitive impairment. Herein we descriptively report these findings for CCMS Waves 1 and 2. Additionally, since the Mini-Mental State Exam (MMSE) was administered at both time points, this investigation sought to compare the relation between MMSE and Scheltens Ratings Scale defined WMHs at both time points in this unique sample of older adults with a mean age of nearly 90 years.



The Cache County Memory Study

The Cache County Memory Study (CCMS) is a population-based study that began in 1994 (1). The county is located in northeastern Utah with the majority of residents living in a large valley, making it convenient for a county-wide population study of dementia (2, 3). Because of the exceptional longevity enjoyed in this area, a focus of the CCMS has been to explore the prevalence and incidence of Alzheimer’s disease (AD) into late old age (1). At the inception of the CCMS investigation, a county-wide attempt, referred to as “Wave 1” ascertainment, was made to contact everyone 65 years and older (N= 5,677 individuals), of which 5,092 elderly residents (90%) became participants in the longitudinal study. As described by Breitner et al. (1) from this initial ascertainment Wave 1 (1995 – 1997) sample, 335 individuals were assessed to have some form of dementia. In this original cohort of participants, using a 0.5 Tesla magnetic resonance imaging (MRI) scanner and standard clinical imaging sequences, studies were obtained on 183 individuals with some level of cognitive impairment or dementia (4) Also as part of the original Wave 1 ascertainment period, 22 individuals assessed to have no dementia, were also imaged as a control sample. Approximately three years later, following a scanner upgrade to 1.5 Tesla a second county-wide ascertainment or “Wave 2” began in 1998 and ran through 2000. In Wave 2 an additional 210 individuals with some form of cognitive impairment or dementia were scanned (3). Combining Waves 1 and 2, over the two ascertainment periods, a total of 415 individuals were scanned, 205 during Wave 1 and 210 during Wave 2, with no overlap in participants from Wave 1 to Wave 2.
Although there are numerous sophisticated neuroimaging methods for quantifying MRI-identified brain pathology, clinical rating remains a standard interpretive method (5). Clinical ratings still have great value because they are easy to perform with minimal time requirements, require only conventional clinical imaging sequences and have no elaborate post-processing requirements (6, 7). Furthermore, studies that have compared clinical ratings with volumetric-based studies or other quantitative methods show reasonable positive correlations (7, 8). A common clinical rating relevant to the aging process and neuropsychological outcome, including degenerative disease and associated neuropsychological impairment has been MRI identification of white matter (WM) abnormalities classified as WM hyperintensities (WMHs) (9). This was previously done for CCMS Wave 1 participants using a simplified rating scale (10), but no ratings have yet been done for the Wave 2 data. One of the most common and widely used clinical rating methods for identifying hyperintense signal abnormalities is the Scheltens Rating Scale (11). This method allows for a standardized assessment with a recognizable metric which then can be used for comparing scans within and between subjects across multiple raters and imaging sessions.
A challenge in performing the Scheltens WMH rating method with the legacy CCMS MRI database is that in addition to the change in scanners and magnetic field strength there were also changes in scan sequences used to assess WM integrity. During the latter part of the 1990’s, the fluid attenuated inversion recovery (FLAIR) sequence (12) became the standard for WM rating, which was the case for Wave 2. Prior to that, WM pathology was defined on the T2 and proton density (PD) sequences, which was the case for Wave 1. The differences in detecting WM pathology between FLAIR and T2/PD based methods are particularly evident when lesion size, location, frequency and volume are compared but less of a problem when just simple detection (i.e., presence/absence) of a WMH is the criterion (13). Given the differences in field strength used during image acquisition between Waves 1 and 2 and that a dual-echo T2/PD sequence was not performed during Wave 2, only WM ratings based on the FLAIR sequence were used for Wave 2.  Comparing PD versus FLAIR in their ability to detect pathology have been reported for a variety of disorders, and while the FLAIR is more sensitive in lesion conspicuity and detection, and especially lesion size (14, 15), differences may not necessarily be present on qualitative rating scales (15). The reason for this is that rating scales are typically ordinal with various classification ranges and rating schemas rather than the interval or ratio scales used in size or volume quantification. With an ordinal rating scale, only a threshold level of pathology is needed for classification rather than an absolute size or volume difference.


Subjects were drawn from the Cache County, Utah, population-based study (1). Table 1 provides total sample and Wave 1 and 2 demographics. Additional, detailed descriptions of the CCMS MRI methods and ascertainment have been published elsewhere (23, 24). In general, as shown in Table 1, age and age ranges, educational levels were similar between Waves 1 and 2 although some differences were statistically present (see Supplementary Table 1-3). Since subjects in Wave 1 were not rescanned at Wave 2, no within-subject comparisons could be made pitting T2/PD to FLAIR.

Table 1. Wave 1 and Wave 2 Sample Demographics

Table 1. Wave 1 and Wave 2 Sample Demographics

Note:  Var = Variables; SD = Standard deviation; Edu = Education; MMSE = Mini-Mental Status Exam.


Measurement instrument

MRI acquisition

In the original Wave 1 cohort, MRI studies were obtained with a 0.5 Tesla field-strength scanner, a clinical imaging standard at that time but at the conclusion of Wave 1 ascertainment there was a change in scanner and magnet field strength. All subsequent images for Wave 2 scanned participants were obtained at 1.5 Tesla. Since this was the only MRI facility in the entire county, there was no alternative imaging center closer than approximately 50 miles away making it impossible to retain a comparable scanner or field strength for Wave 2 neuroimaging.
All subjects underwent MRI with Wave 1 participants scanned on a 0.5 Tesla Phillips MRI scanner while Wave 2 participants were analyzed on a 1.5 Tesla Siemens MRI scanner. Imaging at 0.5 Tesla utilized a quadrature head coil with the following sequence details:   T1-weighted sequence (TR (ms) 500 and TE of 15) with 2 excitations with an acquisition matrix of 256 X 256 with field of view (FOV) at 24 cm, slice thickness of 5mm and gap at 1 mm; axial T2-weighted and PDI sequences (TR (ms) 3148, TE of 31) with 90/1 excitations. Acquisition matrix was 256 X 256 with a FOV of 22 cm, slice thickness of 5mm and a gap of 1.5 mm; coronal dual spin-echo sequence (TR (ms) 3046 and a TE of 30) with 90/1 excitations. Acquisition matrix was 256 X 240 and FOV of 22 cm, slice thickness of 3mm and a gap of 0.3mm.
The 1.5 Tesla scans were obtained using the circular polarized array head coil with the following sequence parameters: Sagittal 1.5 Tesla scans were acquired using a T1 sequence (TR (ms) 500, TE of 14) with 1 excitation with an acquisition matrix of 256 X 192 with a FOV of 22 cm, slice thickness of 5mm, and gap at 1.5 mm. Flip angle was 90; T2, PDI and FLAIR sequences were all obtained in the axial plane, with the T2 sequence (TR (ms) of 6940, TE of 119) with 2 excitations and an acquisition matrix was 512 X 237 with a FOV of 17.5 X 22 cm, slice thickness of 4 mm, and gap at 0.8mm. Flip angle was 170; the PDI sequence (TR (ms) of 3000, TE of 17) with 1 excitation and an acquisition matrix of 256 X 202 with a FOV of 17.9 X 22 cm, slice thickness of 4.5mm, gap at 2 mm  and a flip angle of 170; the FLAIR sequence (TR (ms) of 9000, TE of 104) with 1 excitation with an acquisition matrix of 256 X 163 with a FOV of 17.5 X 22 cm, slice thickness of 5mm, gap at 2 mm and flip angle of 170;  coronal T2-weighted scans (TR (ms) 6940, TE of 119) with 2 excitations with an acquisition matrix of 512 X 237 with a FOV of 16.7 X 21 cm, slice thickness of 5 mm, gap at 1 mm and a flip angle of 170.
For the Wave 1 Scheltens ratings were performed using the T2 and PD sequences. For Wave 2 ratings were based on T2 and FLAIR. Figure 1 depicts two participants, Wave 1 and 2 showing differences in scan parameters and identification of WMHs.

Figure 1. Image differences between Waves 1 and 2 at similar axial levels but with comparable MMSE scores. The top row (T2-weighted) depicts a single participant from Wave 1 (female, age 84) with an MMSE score of 23. The bottom row image (FLAIR) depicts a participant from Wave 2 (participant female, age 98) with an MMSE score of 22. Images were taken axially at the body of the lateral ventricles (left) and centrum semiovale (right).

Figure 1. Image differences between Waves 1 and 2 at similar axial levels but with comparable MMSE scores. The top row (T2-weighted) depicts a single participant from Wave 1 (female, age 84) with an MMSE score of 23. The bottom row image (FLAIR) depicts a participant from Wave 2 (participant female, age 98) with an MMSE score of 22. Images were taken axially at the body of the lateral ventricles (left) and centrum semiovale (right).



Scheltens Ratings Scale

The Scheltens Ratings Scale (11) identifies periventricular, white matter, basal ganglia and infratentorial signal hyperintensities.  The scoring procedures of the original Scheltens Ratings Scale were followed. A training program was initiated with two samples of 10 subjects each from the CCMS data. The first batch of 10 were done for initial training purposes to match an already established expert (EDB) rating. Inter-rater reliability data was not assessed. However, once a 90% or higher accuracy was achieved, the rater moved on to the second sample of 10. Again, each rater had to achieve 90% accuracy before moving on to the entire sample. Each rater was blind to age, sex or diagnosis with each scan independently rated by a minimum of two raters. After rating, if findings were discrepant by more than two rating points per category the raters independently rated the scans a second time, again independently and not knowing where the discrepancy was. If the discrepancy persisted between the two independent raters, EDB arbitrated the final rating. In all cases, the final Scheltens Rating was based on the average of the two independently rated scores per scan.
As stated in the original guidelines for the Scheltens Ratings Scale, it provides “four sum scores in a semi-quantitative way” (11). The four scores, as already mentioned consist of periventricular hyperintensities (PVH), lobar white matter hyperintensities (WMH), basal ganglia hyperintensities (BGH), and infratentorial foci of hyperintensity (IFH). Except for PVH, which were rated from 0-3, all the other scores were rated from 0-6, with higher scores representing a greater degree of WMH.

Mini-Mental State Exam (25)

All participants in the CCMS with suspected cognitive disorders and a panel of healthy controls (N = 22) underwent a full clinical evaluation of dementia.  This evaluation included a clinical history, physical and neurological evaluation, informant-based interviews, standard laboratory chemistry to rule out treatable systemic conditions, and a neuropsychological battery of tests.   Included in the neuropsychological battery was the Mini-Mental State Exam (MMSE) which was uniformly administered to all participants using the standard administration format.

Statistical Analyses

Basic mean differences on Scheltens Ratings between groups were statistically compared with T tests. Pearson product moment correlation (r) were used to assess the relationship between Scheltens Ratings and MMSE scores. We also conducted Factorial Analysis of Variance (ANOVA) examining several variables, including age, education, WMH, BGH, IFH, PVH, Scheltens Total Score, and MMSE score. The effects of sex on these factors was also explored with ANOVA Tukey HSD Test Mean Comparison.
Study procedures were approved by the Institutional Review Boards of Utah State, Duke, and the Johns Hopkins Universities at the time of study enrollment and the archival analysis of the data for the current manuscript was approved by the Institutional Review Board of Brigham Young University.



Scheltens Ratings for total sample across Waves 1 and 2

Table 2 summarizes Scheltens Ratings for all participants across both Waves 1 and 2. Hyperintense signal abnormalities were present in control participants as well as those with cognitive impairment/dementia. As a group, the control sample (mean age = 90.45, S.D. = 5.5) was older than the mean age of either the Wave 1 (mean age = 87.31, S.D. = 6.62) or Wave 2 (mean age = 89.27, S.D. = 5.38) cognitive impairment/dementia samples, although these differences were not statistically significant. In this elderly sample, age was not related to WMH, IFH, BGH or PVH ratings (p>0.05).  Likewise, even though the ‘Healthy Control’ participants had been screened for dementia and not found to meet criteria for any dementing illness, as a group they had WMHs at a level similar and not significantly different from those with dementing illnesses (see Table 2). Removing healthy controls, which were only assessed at Wave 1, Table 3 compares Scheltens Rating by Wave 1 and Wave 2. Wave 1 participants exhibited significantly higher WMH, BGH, PVH and Scheltens Total Score rating than Wave 2 participants.

Table 2. Mean and Standard Deviation of the Scheltens Ratings for All Cases (Wave 1 and 2 Combined)

Table 2. Mean and Standard Deviation of the Scheltens Ratings for All Cases (Wave 1 and 2 Combined)

Note. WMH = White matter hyperintensity; BGH = Basal ganglia hyperintensity; IFH = Infratentorial foci hyperintensity; PVH = Periventricular hyperintensity. No differences were significantly different.

Table 3. Wave 1 and Wave 2 Scheltens Ratings (excluding healthy control cases)

Table 3. Wave 1 and Wave 2 Scheltens Ratings (excluding healthy control cases)

Note. WMH = White matter hyperintensity; BGH = Basal ganglia hyperintensity; IFH = Infratentorial foci hyperintensity; PVH = Periventricular hyperintensity; * p  ≤ .05. ** p ≤ .01. *** p ≤ .001.


Scheltens Rating and MMSE

As shown in Table 4, the expected higher Scheltens Rating scores, indicating increased levels of pathology, were associated with lower MMSE scores. Regardless of whether assessed at Wave 1 or Wave 2, WMH, PVH and Total Scheltens score were significantly correlated with MMSE.  Plots comparing Wave 1 and Wave 2 MMSE performance by WMH and Total Scheltens Ratings are presented in Figure 2. Clearly notable in the scatter plots is that in Wave 1 more participants had lower MMSE scores, suggesting that overall as a group Wave 1 participants had greater levels of cognitive impairment.  In support of that supposition, the group mean MMSE score at Wave 1 was significantly lower than the group MMSE mean at Wave 2 (see Table 2 Supplementary Table).

Figure 2. Plots comparing Wave 1 (left) and Wave 2 (right) Mini-Mental Status Exam (MMSE) performance by white matter hyperintensities (WMH) and Total Scheltens Rating

Figure 2. Plots comparing Wave 1 (left) and Wave 2 (right) Mini-Mental Status Exam (MMSE) performance by white matter hyperintensities (WMH) and Total Scheltens Rating

Note: WMH = White matter hyperintensity; MMSE = Mini Mental Status Exam

Table 4. Correlation between Scheltens Ratings and Mini-Mental Status Exam (MMSE)

Table 4. Correlation between Scheltens Ratings and Mini-Mental Status Exam (MMSE)

Note. WMH = White matter hyperintensity; BGH = Basal ganglia hyperintensity; IFH = infratentorial foci hyperintensity; PVH = Periventricular hyperintensity; MMSE = Mini-Mental Status Exam. * p  ≤ .05. ** p ≤ .01. *** p ≤ .001.



As noted, both controls and those with dementia displayed WMH, although age was not statistically associated with WNH in this sample. However, higher Scheltens Rating scores were statistically correlated with MMSE performance. Since the CCMS represents one of the major population-based longitudinal studies of advanced aging, we felt that it was important to better characterize white matter pathology in this cohort because of the important role that such pathology plays in aging and neurodegenerative disease (9). As expected, in this elderly CCMS sample increased levels of WM pathology as reflected in high Scheltens Ratings were associated with worse MMSE performance, although less robust in the Wave 2 cohort. The reduced association at Wave 2 likely reflects the differences in degree of cognitive impairment between the two waves of ascertainment. Since Wave 1 was the initial county-wide assessment of dementia prevalence in the population, it captured all individuals with cognitive disorders from those with mild dementia to those who were very functionally impaired. By contrast, Wave 2 was designed to capture new incident cases of dementia occurring over an approximately three-year interval. Consequently, by design, incident dementia arising in the population within the three-year window would have included milder cases of cognitive disorder. Indeed, in review of Table 1, overall MMSE score was significantly lower in Wave 1.
Other studies that have used the Scheltens Ratings Scale (8, 26, 27) have found similar results where WM burden as reflected in the overall amount of hyperintense signal abnormality was significantly associated with reduced cognitive performance. Furthermore, as outlined by Tschanz et al. (22) the CCMS study is providing novel insights into the genetics, psychosocial and environmental risk factors of AD, late-life cognitive decline and the clinical progression of dementia after onset. The current legacy CCMS WM Scheltens Ratings will permit further investigation of the role of white matter pathology and clinical correlates, including neuropsychological, relation to neuropsychiatric symptoms and presentation along with functional impairment.
Wave 1 also was the time period where 22 control subjects without dementia at the time scanned were examined. As a group, these healthy yet very elderly controls that averaged 90 years in age did have high Scheltens Rating Scale scores that differed minimally and non-significantly from Wave 1 and 2 participants who had some cognitive impairment or dementing illness, in particular, Wave 2 participants. As reflected in the control sample, in the presence of normal MMSE findings, WMHs do not portend cognitive impairment (28, 29). However, WM pathology and its severity do show a relationship in those with dementing illness where greater WM abnormalities are associated with lower MMSE scores.
These findings are important given that recent studies of white matter integrity measured with visual rating scales differentiate clinical groups, including vascular dementia and Alzheimer’s disease, where deep white matter hyperintensities increase risk of vascular dementia (21). Additional research also demonstrates that visual rating of white matter lesions among older adults with cerebrovascular disease increases the risk of progression from mild cognitive impairment to dementia (3.5 to 3.8-fold increase) (20).
While multiple studies have examined the association of white matter integrity and cognition, few have reported specifically on MMSE. The results of the current study are important because of the unique population of older adults that make up the Cache County data. Data from the Rotterdam Scan Study (30) examined change in MMSE scores from baseline to follow-up (mean interval 3.4 years) in a prospective study of older adults, and examined this change in association with visual rating scores of white matter lesions. Consistent with the present study, van Dijk and colleagues demonstrated that periventricular white matter lesions were associated with a decline in MMSE across the study period. Similarly, Defrancesco and colleagues (31) examined the influence of neuropsychological tests, including the MMSE, and white matter lesion burden on the conversion of mild cognitive impairment to dementia. They demonstrated that visual ratings of white matter lesion burden are predictive of orientation from the MMSE and total MMSE scores, and that increase in white matter burden over time was associated with a decline in MMSE and performance on an object naming test. Additionally, individuals in this study who converted from mild cognitive impairment to dementia had higher white matter lesion burden, particularly in periventricular regions, on visual rating scales compared to non-converters. While not with a visual rating system, Li et al. (32) similarly demonstrated that mean diffusivity of the corpus callosum is negatively correlated with performance on the MMSE, but only among older adults with amnestic mild cognitive impairment – multidomain subtype. The current findings offer supportive data of the association between WM burden and general cognitive functioning in older adults with dementia.
There are several prominent limitations of the current study. In samples younger than the current sample of Cache County participants, MMSE may not be impacted by white matter change. For example, in a study of middle-aged adults (mean age between 59-61), MMSE scores were not related to periventricular hyperintensities nor deep white matter hyperintensities (33). As such, the effects of WM burden on MMSE scores may be mediated by age or other factors common in older populations. It is also possible that while WM burden can impact some aspects of cognition in younger adults (e.g., processing speed), the MMSE and similar tests with low ceiling effects, are simply not sensitive enough in younger adults.
In addition, the noted white matter pathology and associated clinical ratings are putatively related to cerebrovascular disease. That is, as cerebrovascular disease increases, microvascular ischemic injury occurs, resulting in vascular cognitive impairment. However, research on the MMSE versus other mental status examinations suggest that other measures may be more sensitive to vascular pathology. For example, the Montreal Cognitive Assessment (MoCA) is one such measure that demonstrates superior sensitivity to cognitive change of a vascular etiology (35). As such, the use of the MMSE, which was the gold standard at the time of data collection, is limited relative to other possible measures.
The current findings are also considered in the context of recent literature indicating that the impact of white matter burden on cognition may also be mediated by other pathological factors. Specifically, the relative importance of white matter lesions in cognitive functioning was called into question in a recent study by Claus et al. (34) who demonstrated that while white matter lesions do predict non-memory cognitive performances, the amount of variance accounted for in memory performance was largely mediated by medial temporal atrophy, not white matter integrity. They further argued that non-memory performances are related to white matter only when medial temporal atrophy is present, and even then, white matter lesions only account for a small amount of the variance. The presence of medial temporal abnormalities was not assessed in the current study.
The difference in field strength and different MR platforms represents a major limitation. Nonetheless, generally comparable findings whether using 0.5 Tesla or 1.5 provided a method for qualitative ratings that related to cognitive (MMSE) status. Since image acquisition was not compatible with any current methods for automated image quantification, the current findings cannot be related to other aspects of contemporary imaging used to quantify white matter integrity, like diffusion tensor imaging. The control sample was small and not well stratified by age.  Now that these WM ratings have been completed for the CCMS study, this will permit investigation of the role of WM pathology in healthy aging and dementia including pre-morbid health and lifestyle where neuroimaging factors have not been explored (see 3 & 22). This study demonstrates that the Scheltens Rating System can be effectively applied to an older community-based population sample using different scanners and image platforms. The ease with which the Scheltens Ratings can be made and the fact that they relate to cognitive outcome adds to the literature on the use of white matter pathology in aging and dementia.  A recent review on WMH, cognitive, and dementia provides an update on the relationship between WMH and cognitive decline and provides support for continued investigation of clinical ratings (9). While visual-based clinical rating scales may not be the best for measuring progression of white matter pathology, their ease and universal applicability are well suited for clinical use.  Improved understanding of the clinical significance of white matter findings in the process of aging and neurodegenerative disease, especially how more elaborate neuroimaging methods quantify white matter pathology and relate to clinical ratings, will likely add to the utility of these simple rating methods that can be applied to any individual who has undergone an MRI.


Funding & Acknowledgements: This work was supported by a Family, Home and Social Science College grant from BYU and NIA grant R01AG11380

Conflict of interests: No conflicts to report.



1.    Breitner, J.C., et al., APOE-epsilon4 count predicts age when prevalence of AD increases, then declines: The Cache County Study. Neurology, 1999. 53(2): p. 321-31.
2.    Norton, M.C., et al., Telephone adaptation of the Modified Mini-Mental State Exam (3MS). The Cache County Study. Neuropsychiatry Neuropsychol Behav Neurol, 1999. 12(4): p. 270-6.
3.    Tschanz, J.T., et al., Stressful life events and cognitive decline in late life: moderation by education and age. The Cache County Study. Int J Geriatr Psychiatry, 2013. 28(8): p. 821-30.
4.    Bigler, E.D., et al., Dementia, quantitative neuroimaging, and apolipoprotein E genotype. AJNR Am J Neuroradiol, 2000. 21(10): p. 1857-68.
5.    Meijs, A.P., et al., How does additional diagnostic testing influence the initial diagnosis in patients with cognitive complaints in a memory clinic setting? Age Ageing, 2015. 44(1): p. 72-7.
6.    Frisoni, G.B., et al., The clinical use of structural MRI in Alzheimer disease. Nat Rev Neurol, 2010. 6(2): p. 67-77.
7.    Kim, G.H., et al., T1-weighted axial visual rating scale for an assessment of medial temporal atrophy in Alzheimer’s disease. J Alzheimers Dis, 2014. 41(1): p. 169-78.
8.    Wang, L., et al., White matter hyperintensities and changes in white matter integrity in patients with Alzheimer’s disease. Neuroradiology, 2011. 53(5): p. 373-81.
9.    Prins, N.D. and P. Scheltens, White matter hyperintensities, cognitive impairment and dementia: an update. Nat Rev Neurol, 2015. 11(3): p. 157-65.
10.    Bigler, E.D., et al., Role of white matter lesions, cerebral atrophy, and APOE on cognition in older persons with and without dementia: the Cache County, Utah, study of memory and aging. Neuropsychology, 2003. 17(3): p. 339-52.
11.    Scheltens, P., et al., A semiquantative rating scale for the assessment of signal hyperintensities on magnetic resonance imaging. J Neurol Sci, 1993. 114(1): p. 7-12.
12.    De Coene, B., et al., MR of the brain using fluid-attenuated inversion recovery (FLAIR) pulse sequences. AJNR Am J Neuroradiol, 1992. 13(6): p. 1555-64.
13.    Tourbah, A., et al., Magnetic resonance imaging using FLAIR pulse sequence in white matter diseases. J Neuroradiol, 1996. 23(4): p. 217-22.
14.    Paniagua Bravo, A., et al., A comparative MRI study for white matter hyperintensities detection: 2D-FLAIR, FSE PD 2D, 3D-FLAIR and FLAIR MIP. Br J Radiol, 2014. 87(1035): p. 20130360.
15.    Sibbitt, W.L., Jr., et al., Fluid Attenuated Inversion Recovery (FLAIR) imaging in neuropsychiatric systemic lupus erythematosus. J Rheumatol, 2003. 30(9): p. 1983-9.
16.    Scheltens, P., et al., White matter changes on CT and MRI: an overview of visual rating scales. European Task Force on Age-Related White Matter Changes. Eur Neurol, 1998. 39(2): p. 80-9.
17.    Otsuka, Y., et al., Diffuse tract damage in the hemispheric deep white matter may correlate with global cognitive impairment and callosal atrophy in patients with extensive leukoaraiosis. AJNR Am J Neuroradiol, 2012. 33(4): p. 726-32.
18.    Quinque, E.M., et al., Identifying the neural correlates of executive functions in early cerebral microangiopathy: a combined VBM and DTI study. J Cereb Blood Flow Metab, 2012. 32(10): p. 1869-78.
19.    Vernooij, M.W., et al., White matter microstructural integrity and cognitive function in a general elderly population. Arch Gen Psychiatry, 2009. 66(5): p. 545-53.
20.    Clerici, F., et al., Does vascular burden contribute to the progression of mild cognitive impairment to dementia? Dement Geriatr Cogn Disord, 2012. 34(3-4): p. 235-43.
21.    Smith, C.D., et al., Peripheral (deep) but not periventricular MRI white matter hyperintensities are increased in clinical vascular dementia compared to Alzheimer’s disease. Brain Behav, 2016. 6(3): p. e00438.
22.    Tschanz, J.T., et al., The Cache County Study on Memory in Aging: factors affecting risk of Alzheimer’s disease and its progression after onset. Int Rev Psychiatry, 2013. 25(6): p. 673-85.
23.    Bigler, E.D., et al., White matter lesions, quantitative magnetic resonance imaging, and dementia. Alzheimer Dis Assoc Disord, 2002. 16(3): p. 161-70.
24.    Tschanz, J.T., et al., Dementia diagnoses from clinical and neuropsychological data compared: the Cache County study. Neurology, 2000. 54(6): p. 1290-6.
25.    Folstein, M.F., S.E. Folstein, and P.R. McHugh, «Mini-mental state». A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res, 1975. 12(3): p. 189-98.
26.    Brickman, A.M., et al., Measuring cerebral atrophy and white matter hyperintensity burden to predict the rate of cognitive decline in Alzheimer disease. Arch Neurol, 2008. 65(9): p. 1202-8.
27.    Prasad, K., et al., White matter disease independently predicts progression from mild cognitive impairment to Alzheimer’s disease in a clinic cohort. Dement Geriatr Cogn Disord, 2011. 31(6): p. 431-4.
28.    Bennett, I.J., et al., Age-related differences in multiple measures of white matter integrity: A diffusion tensor imaging study of healthy aging. Hum Brain Mapp, 2010. 31(3): p. 378-90.
29.    Kerchner, G.A., et al., Cognitive processing speed in older adults: relationship with white matter integrity. PLoS One, 2012. 7(11): p. e50425.
30.    van Dijk, E.J., et al., Progression of cerebral small vessel disease in relation to risk factors and cognitive consequences: Rotterdam Scan study. Stroke, 2008. 39(10): p. 2712-9.
31.    Defrancesco, M., et al., Impact of white matter lesions and cognitive deficits on conversion from mild cognitive impairment to Alzheimer’s disease. J Alzheimers Dis, 2013. 34(3): p. 665-72.
32.    Li, H., et al., Different patterns of white matter disruption among amnestic mild cognitive impairment subtypes: relationship with neuropsychological performance. J Alzheimers Dis, 2013. 36(2): p. 365-76.
33.    Soriano-Raya, J.J., et al., Tract-specific fractional anisotropy predicts cognitive outcome in a community sample of middle-aged participants with white matter lesions. J Cereb Blood Flow Metab, 2014. 34(5): p. 861-9.
34.    Claus, J.J., et al., Cerebral White Matter Lesions have Low Impact on Cognitive Function in a Large Elderly Memory Clinic Population. J Alzheimers Dis, 2018. 63(3): p. 1129-1139.
35.    Van Heugten, C. M., Walton, L., & Hentschel, U. Can we forget the Mini-Mental State Examination? A systematic review of the validity of cognitive screening instruments within one month after stroke. Clinical rehabilitation, 2015;29(7), 694-704.



C. Moran1, A. Scotto di Palumbo2, J. Bramham1, A. Moran1, B. Rooney1, G. De Vito2, B. Egan2


1. School of Psychology, University College Dublin, Dublin, Ireland; 2. School of Public Health, Physiotherapy and Sports Science, University College Dublin, Dublin, Ireland.

Corresponding Author: Catherine Moran, School of Psychology, Trinity College Institute for Neuroscience, Trinity College Dublin, College Green, Dublin 2, Ireland, E-mail:

J Prev Alz Dis 2017 inpress
Published online February 20, 2018,



Objectives: To investigate the impact of a six-month multi-ingredient nutrition supplement intervention (Smartfish®), containing omega-3 polyunsaturated fatty acids (PUFAs), vitamin D, resveratrol, and whey protein, on cognitive function in Irish older adults.
Design: Double-blind, randomised controlled trial ( NCT02001831). A quantitative, mixed-model design was employed in which the dependent variable (cognitive function) was analysed with a between-subjects factor of group (placebo, intervention) and within-subjects factor of testing occasion (baseline, three-months, six-months).
Setting: Community-based intervention including assessments conducted at University College Dublin, Ireland.
Participants: Thirty-seven community-dwelling older adults (68-83 years; mean (x̄)= 75.14 years; standard deviation (SD)= 3.64; 18 males) with normal cognitive function (>24 on the Mini Mental State Examination) were assigned to the placebo (n= 17) or intervention (n= 20) via a block randomisation procedure.
Intervention: Daily consumption for six-months of a 200mL liquid juice intervention comprising 3000mg omega-3 PUFAs [1500mg docosahexaenoic acid (DHA) and 1500mg eicosapentaenoic acid (EPA)], 10μg vitamin D3, 150mg resveratrol and 8g whey protein isolate. The placebo contained 200mL juice only.
Measurements: A standardised cognitive assessment battery was conducted at baseline and follow-ups. Individual test scores were z-transformed to generate composite scores grouped into cognitive domains: executive function, memory, attention and sensorimotor speed. Motor imagery accuracy and subjective awareness of cognitive failures variables were computed from raw scores.
Results: A hierarchical statistical approach was used to analyse the data; first, by examining overall cognitive function, then by domain, and then by individual test scores. Using mixed between-within subjects, analyses of variance (ANOVAs), no significant differences in overall cognitive function or composite cognitive domains were observed between groups over time. The only significant interaction was for Stroop Color-Word Time (p< 0.05). The intervention group demonstrated reduced task completion time at three- and six-month follow-ups, indicating enhanced performance. Conclusion: The present nutrition intervention encompassed a multi-ingredient approach targeted towards improving cognitive function, but overall had only a limited beneficial impact in the older adult sample investigated. Future investigations should seek to establish any potential clinical applications of such targeted interventions with longer durations of supplementation, or in populations with defined cognitive deficits.

Key words: Cognitive failures, executive function, aging, nutrition, supplementation.

Abbreviations and Symbols: ANOVA: Analysis of Variance; AVLT: Auditory Verbal Learning Test; BMI: Body Mass Index; CFQ: Cognitive Failures Questionnaire; COWA: Controlled Oral Word Association; C-W: Color-Word; DHA: Docosahexaenoic acid; EPA: Eicosapentaenoic acid; INT: Intervention Group; MI: Motor Imagery; MMSE: Mini Mental State Examination; PI: Principal Investigator; PLAC: Placebo Group; PUFA: Polyunsaturated fatty acid; RCT: Randomised controlled trial; SD: Standard deviation; TMT: Trail Making Test; TUG: Timed Up and Go; UCD: University College Dublin; WAIS-III: Wechsler Adult Intelligence Scale III; x̄: Mean.




Cognitive function tends to decline with advancing age. Older adults may experience compromises in memory, attention and executive functioning that significantly impair their capacity to cope with daily social and occupational demands (1). In the quest to understand possible mechanisms, recent research has explored the role of modifiable risk factors, such as physical activity (2) and diet (3), in curbing age-related cognitive decline. Of the dietary factors investigated to date, omega-3 polyunsaturated fatty acids (PUFAs) has the highest evidence-based potential for clinical use (4). The precise nature of this impact, however, remains unclear. To illustrate, in some studies, high omega-3 PUFA consumption is associated with improved cognitive functioning or reduced risk of dementia; whereas in others, no such effect is evident (5-9). A Cochrane review (3) reported on three randomised controlled trials (RCTs) (10-12) in this field and found no benefit of omega-3 PUFA supplementation on cognitive function in healthy elderly. However, more recent RCTs have demonstrated enhanced executive functioning (13) and object location memory task performance (although no effect on the Auditory Verbal Learning Test; AVLT) (14) after omega-3 PUFA supplementation in healthy older adults. Vitamin D insufficiency has been suggested as a potential modifiable risk for age-associated cognitive decline (15, 16). In this regard, two prospective population-based cohort studies (17, 18) examined this association, using the Mini Mental State Examination (MMSE) (19) and at least one version of the Trail Making Test (TMT) (20) in older adults at baseline and follow-up. Again, inconsistency of findings is apparent; whereas poorer cognitive function exists in participants who are vitamin D deficient (17), negligible evidence of a link between vitamin D and executive function or incident cognitive decline has also been observed (18). In addition, a 12-year population-based longitudinal study of 1058 adults (aged >50 years at baseline) found an association between vitamin D deficiency and poorer performance on a range of baseline cognitive assessments, but no association between vitamin D status and task performance or cognitive decline at follow-ups (21). As such, RCTs are warranted to causally determine the benefits, if any, of vitamin D supplementation in the treatment or prevention of cognitive decline.
Emerging research suggests that resveratrol, a polyphenol plant compound, may modulate mechanisms of neuronal aging (22-24). However, the complexity of the biological substrates of polyphenols in cells and animals represents a major challenge in extending this research to humans (25). In this regard, human studies evaluating the role of resveratrol on cognitive function are scant. The beneficial role of whey protein supplementation has also been examined; mostly regarding physiological health outcomes, including enhanced muscle mass (26), increased artery elasticity and decreased risk of heart disease and stroke (27). Despite the significant positive associations between these outcomes and brain function, interventional evidence is lacking on the specific role of dairy constituents in neurocognitive health over the lifespan (28).
In summary, evidence concerning the benefits of nutrition supplementation on cognitive processes in older adults remains inconclusive. Moreover, previous research has focused almost exclusively on the impact of individual ingredients on cognitive function. Against this background, the present study addresses this gap in the literature by experimentally evaluating a six-month multi-ingredient supplement intervention containing omega-3 PUFAs, vitamin D, resveratrol and whey protein on cognitive function in Irish older adults. It was hypothesised that the experimental intervention would improve overall cognitive functioning, executive function, memory, attention, sensorimotor speed, motor imagery (MI) accuracy and subjective awareness of cognitive failures, compared to the placebo condition.




A double-blind RCT was employed to investigate the efficacy of a six-month, multi-ingredient nutrition supplement intervention for improving cognitive functioning in older adults; specifically, effects on executive function, memory, attention, sensorimotor speed, MI accuracy, and subjective awareness of cognitive failures were assessed. For this quantitative, mixed-model design, the dependent variable (“cognitive function”) was analysed with respect to a between-subjects factor of “group allocation” (placebo or intervention group) and a within-subjects factor of “testing occasion” (baseline, three-months, and six-months).

Ethical approval

All study procedures were enacted in accordance with the ethical codes of conduct of the Psychological Society of Ireland and the guidelines of the Declaration of Helsinki (2008, 2013). The research protocol was reviewed under the broader Smartfish® project and granted ethical approval from the University College Dublin (UCD) Human Research Ethics-Sciences Board (reference: LS-13-28-Egan). Participants provided written informed consent prior to study enrolment. No animals were included in this research.

Sample size calculation and study power

To calculate an estimate for sample size, an alpha value of 0.05 and beta value of 0.2 was set to ensure Power would be 0.8. Given our two allocation groups and three testing occasions, this determined that a sample size of 28 participants would be required to detect a medium effect size (f= 0.25) (GPower v3.1). To account for potential drop-out rate, we aimed to recruit more participants prior to randomisation. A post-hoc calculation of our actual power based on 37 trial-completers was conducted and demonstrated a 0.914 power to detect medium effects, in line with our intended goal.


Participants were recruited via a combination of methods including an advertisement placed in a national newspaper (Irish Times), invitations issued on the UCD alumni website, and recruitment flyers distributed to local elderly organisations and retirement homes. Individuals who expressed interest in the study were invited to UCD and provided with an information leaflet, which addressed issues of confidentiality, anonymity and data protection. At this point, a consent form was signed in the presence of the researcher. Eligibility for participation was then established from a pre-screening examination with a medical doctor. Participants aged 65 years or over, defined as ‘healthy’ (disease free) (29), who were independent, mobile and capable of completing the trial, and who scored above 24 on the MMSE [19] were considered eligible. Potential participants who concurrently fulfilled these inclusion prerequisites, and did not report current or recent (8-week) use of fish oil, or vitamin D or whey protein supplements, were subsequently selected for the trial.
Only participants who completed assessments at all three time-points were included in the statistical analysis (per protocol analysis). The total sample (N= 37) comprised 18 males and 19 females with an overall mean age of 75.14 years (SD= 3.64; range 68- 83 years). Of the 37 ‘trial completers’, 17 had been randomised into the placebo group (PLAC) and 20 had been randomised into the intervention group (INT) (see Figure 1 for a consort diagram detailing study participation). The principal investigator (PI), blind to the assessments, conducted this random allocation procedure by means of a block randomisation. Envelopes were selected from an opaque container, which contained an equal distribution of placebo and supplement. Once an envelope had been drawn it was not returned prior to the subsequent randomisation.


In the active arm, the intervention liquid nutrient support (quantity 200mL per day; energy 200kcal per day) comprised 3000mg of long-chain omega-3 PUFAs [as 1500mg docosahexaenoic acid (DHA) and1500mg eicosapentaenoic acid (EPA)], 10μg of vitamin D3, 150mg of resveratrol, and 8g of whey protein isolate. The placebo nutrient support contained 200mL of juice only (energy 100kcal per day). Smartfish®, a Norwegian biotech company, provided both the supplement and placebo as ready-to-drink, palatable, pomegranate and apple flavoured juice formulations, presented in identically sealed TetraPak cartons. The formulations were indistinguishable in appearance and taste, and participants were required to consume their allotted formulation daily for a period of six-months. Research staff and participants were completely blind to group allocation until completion of the data collection. Participants received the juice cartons immediately following their baseline assessment, and these were replenished following their intermediate assessment. Compliance to the supplementation protocol was recorded using a daily tick-box diary completed by each participant.

Cognitive assessments

Between October 2013 and January 2015 data were collected in the Human Performance laboratory at the UCD Institute for Sport and Health. An extensive cognitive assessment battery comprising seven measures was conducted at three time points [baseline, intermediate (three-months) and follow-up (six-months)]. The battery was an English replication of that used in a previous investigation of omega-3 PUFA supplementation and cognitive function (13) with the addition of the Timed Up and Go (TUG) test (30) and the Cognitive Failures Questionnaire (CFQ) (31). The Trail Making Test (TMT) [20] was administered in two parts: Part A assessed sensorimotor speed and visual tracking and part B measured cognitive flexibility. The Auditory Verbal Learning Test (AVLT) (32) examined learning (immediate recall), retention (30-minute delayed recall) and retrieval (30-minute delayed recognition) of newly acquired verbal information. Alternate versions of the AVLT were used at follow-ups to prevent practice effects. The Stroop test (33) was administered as a measure of selective attention, processing speed, and susceptibility to cognitive interference. The version used consisted of two components, namely Color and Color-Word (C-W) tasks (34). The Controlled Oral Word Association test (COWA) (35) measured executive functioning and was administered in two parts to explore phonemic and categorical verbal fluency. The Digit Span test, taken from the Wechsler Adult Intelligence Scale III (WAIS-III) (36), comprised two different tests; the digits forward task, which measured attentional capacity and digits backward, which assessed working memory performance.
The TUG (30) is a chronometric task designed to measure MI accuracy. MI is the mental simulation of an action in the absence of execution (37). The standard version of the task (TUG Real) (38) measures, in seconds, the time taken for participants to stand from a standard chair, walk a distance of three metres, turn around, return to the chair and sit down again. In the MI task version (TUG Imagined), participants perform this task in their imagination and then, signal ‘stop’ to terminate the task. This measure was added to the battery as recent research has focused on the interface between mental and physical functioning, namely MI, as a potential biomarker of cognitive decline (39). Finally, the CFQ (31) is a 25-item self-report inventory that measures cognitive lapses in everyday life. It assesses frequencies of self-reported anomalies in perception, memory and motor function over the previous month. This aspect of cognitive function is often neglected in the literature, which focuses almost exclusively on subtle changes in performance as assessed by objective, lab-based measures. Few studies investigate the relative impact on real day-to-day functioning; and as impaired meta-awareness of cognitive failures has been demonstrated in early neurological conditions (40), this measure was included in the present study to fully establish the clinical utility of the intervention.
Data collection for each of the three testing occasions lasted approximately 45-minutes and was conducted by trained psychology Research Assistants under the supervision of a Clinical Neuropsychologist using scripted instructions and following standardised procedures. Each participant was issued a unique subject number at study entry. To ensure anonymity, only this subject number was used on the data recording forms; no other identifying information was linked to the assessments. Testing was conducted in the same quiet room at approximately the same time in the morning. Consumption of coffee and tea was not permitted before or during testing; participants were provided with a standard breakfast prior to commencement.

Statistical Analysis

IBM SPSS Statistics 20 (41) was used to analyse the data. Preliminary analyses were conducted to compare the placebo and intervention groups on demographics and baseline cognitive function variables (see Tables 1 and 2). Independent samples t-tests were used to compare the groups on continuous variables such as age, height, body mass, body mass index (BMI), number of years of full-time education and baseline cognitive test variables; while chi square tests compared the groups for gender and categorisation of highest qualification.

Table 1. Descriptive and Test Statistics Comparing Baseline Group Characteristics

Table 1. Descriptive and Test Statistics Comparing Baseline Group Characteristics

Note. Abbreviations: INT: Intervention group; PLAC: Placebo group. Gender data expressed as: n male (n female). Age, Height, Body Mass, BMI and Education data expressed as mean (SD). Qualification data expressed as frequency counts.

Table 2. Independent Samples t-Tests Comparing Baseline Cognitive Function Across Groups

Table 2. Independent Samples t-Tests Comparing Baseline Cognitive Function Across Groups

Note. Abbreviations: AVLT: Auditory Verbal Learning Test; CFQ: Cognitive Failures Questionnaire; INT: Intervention group; PLAC: Placebo group; SD: standard deviation; Stroop C-W: Stroop Color-Word Time; TMT: Trail Making Test; TUG: Timed Up and Go.



Following the protocol of previous research (12, 13), individual cognitive test scores were z-transformed and averaged to generate composite scores for each time point that were grouped for analysis in the following cognitive domains:
Executive function: [Z Phonemic Total + Z Category Total – Z TMT (part B-part A)/part A – Z Stroop (part C-W – part C)]/4
Memory: (ZAVLT Total + Z15 AVLT Delay + Z15 AVLT Recognition + ZDigit Span Backward)/4
Attention: ZDigit Span Forward
Sensorimotor speed: (-ZTMT A Time – ZStroop part C – ZStroop part C-W)/3
To establish a measure of MI accuracy that allowed for comparisons between groups, participants’ durations when performing the TUG Real and TUG Imagined were entered in the following formula, yielding an objective index, namely ‘TUG Delta’ (30):
TUG Delta: [(TUGr – TUGi)/(TUGr + TUGi)/2]*100

Finally, the subjective awareness of cognitive failures variable comprised raw CFQ total scores.
Subsequently, a three-tier hierarchical approach was adopted to test the research hypotheses for a Group X Time interaction as evidence of change due to the intervention (see Table 3). Firstly, a mixed between-within subjects ANOVA investigated whether there was an effect of intervention on overall cognition at six-months using composite variables. The alpha coefficient used as the significance criterion was 0.05. Secondly, each composite variable (executive function, memory, attention, sensorimotor speed, MI accuracy, subjective awareness of cognitive failures) was explored separately using a number of individual mixed between-within subjects ANOVAs. Thirdly, each individual test variable was investigated for an effect of intervention compared to placebo using mixed between-within subjects ANOVAs.
Sensitivity analyses using intention-to-treat methods for dealing with dropout-missing data (last observation carried forward, imputing means of the group, imputing means of the other group) were also conducted, and the inferential analyses repeated. However, as there were no major differences in findings between methods, only the results of the per-protocol analysis of 37 ‘trial completers’ are reported here.



Cognitive function data for 37 participants, excluding the 14 dropout participants (27.45%), were available after six-months of the intervention. Seven participants withdrew from the trial before their intermediate assessment (1 male, 6 females; mean age 77.00 ± 5.60 years), and a further 7 withdrew before their final assessment (1 male, 6 females; mean age 73.86 ± 4.45 years). See Figure 1 for more detail on participant recruitment and retention. Chi-square and independent t-test analyses demonstrated that ‘excluded’ participants were not significantly different from ‘included’ participants regarding demographic characteristics or baseline cognitive function.

Figure 1. Consort diagram detailing study participation

Figure 1. Consort diagram detailing study participation


Data from the 37 trial completers were inspected for outliers using boxplots and any data points that extended above or below two standard deviations from the mean were excluded from further analysis. In total, 1.98% of data points were excluded as outliers and a further 0.66% of data points were counted as missing.
Continuous variables approximated normal distributions; thus, parametric statistics were utilised. At baseline, groups (PLAC, INT) were matched on demographic characteristics and cognitive function (Tables 1 and 2). Compliance to the supplementation protocol, using the self-report daily tick-box diary, was 95±5% for PLAC and 96±4% for INT.
Using mixed between-within subjects ANOVAs, no statistically significant differences in overall cognitive function or cognitive function domain scores (executive function, memory, attention, sensorimotor speed, MI accuracy and subjective awareness of cognitive failures) were observed for either group over six-months (Table 3). There was no evidence to suggest that the groups differed; that is, there was no difference in the efficacy of the intervention compared with the placebo on these cognitive variables. The effect of time, regardless of group, was significant for overall cognitive function, executive function and memory. Inspection of the z-scored means demonstrated that participants improved on these variables over the multiple testing occasions.


Table 3. Mixed Between-Within Subjects ANOVAs for Composite and Individual Test Variables

Table 3. Mixed Between-Within Subjects ANOVAs for Composite and Individual Test Variables

Note. 0, 3 and 6 denote month of testing. Abbreviations: AVLT: Auditory Verbal Learning Test; INT: Intervention group; MI: Motor Imagery; PLAC: Placebo group; SD: standard deviation; Stroop C-W: Stroop Color-Word Time; TMT: Trail Making Test; TUG: Timed Up and Go; *Significance at the .05 level; **Significance at the .01 level; ***Significance at the .001 level; a. Calculated from the formula: [Z Phonemic Total + Z Category Total – Z TMT (part B-part A)/part A – Z Stroop (part C-W – part C)]/4. The resulting data for ‘executive function’ are based on z-scores; b. Calculated from the formula: (ZAVLT Total + Z15 AVLT Delay + Z15 AVLT Recognition + ZDigit Span Backward)/4. The resulting data for ‘memory’ are based on z-scores; c. Calculated from the formula: Zdigit span-forward. The resulting data for ‘attention’ are based on z-scores; d. Calculated from the formula: (-ZTMT A Time – ZStroop part C – ZStroop part C-W)/3. The resulting data for ‘sensorimotor speed’ are based on z-scores; e. Calculated from the formula: [(“TUG” – “TUGi”)/(“TUG” + “TUGi”/2] x 100; f. Calculated from CFQ total scores; g. Significant effect of time for the intervention group, variance ratio F (2, 58) = 8.48; probability (p) <.05;  No significant effect of time for the control group, F (2, 58) = 1.21; p >.05;  No significant group effect at baseline, F (1, 58) = 0.12; p >.05;  Significant group effect at 3-months, F (1, 58) = 14.57; p <.01;  Significant group effect at 6-months, F (1, 58) = 6.53; p <.05.



This analytic procedure was repeated for all the individual cognitive test variables (Table 3). The only significant interaction between group and time was for ‘Stroop Color-Word Time’. However, it should be noted that a Bonferroni adjustment would remove this effect. Tests of simple effects were conducted to explore the nature of this interaction (see ‘Notes’, Table 3 for exact statistics). Results revealed a significant effect of time (reduction in scores) for the intervention group; no such significant effect was observed in the control group. The tests revealed that groups did not significantly differ at baseline, but by 3-months, the intervention group demonstrated significantly lower time scores than the control group. These effects were also evident at 6-months. This suggests that Stroop Color-Word performance improved over time for the intervention group compared to the placebo group.
No other significant interactions or group effects occurred. However, significant effects for time were observed for TMT A Time, Stroop Color Time and AVLT Delay variables. Using tests of within-subjects contrasts these effects were observed to be linear. Irrespective of group, participants showed a pattern of dis-improvement on the Stroop Color Time and AVLT Delay variables, and a pattern of performance improvement on the TMT A Time, task over the three testing occasions.



The present study investigated the effects of a six-month multi-ingredient nutrition supplement intervention on cognitive function in community-dwelling Irish older adults. Although some previous research has demonstrated beneficial effects of individual ingredients on cognitive function in interventions with nutrition supplementation, the present study employed a novel multi-ingredient approach with nutrients combined to target cognitive function in older adults. Importantly, the assessment of cognitive function was comprehensive, with only two previous studies in which a comparable range of cognitive outcomes was examined (12, 13). Overall, no statistically significant differences in cognitive functioning or in composite cognitive outcomes were observed between groups over time. Therefore, the hypotheses stating that overall cognitive function, executive function, memory, attention, sensorimotor speed, MI accuracy and subjective awareness of cognitive failures would improve in the intervention group compared to placebo group at six-months were not supported. However, with one exception, Stroop Color-Word performance did improve for participants receiving the intervention compared to the placebo at three- and six-month follow ups. However, it should be noted that a Bonferroni adjustment would remove this effect. Thus, the multi-ingredient nutrition intervention had only limited beneficial impact on cognitive functioning after six-months of supplementation in an Irish, community-dwelling older adult population.
When looking to studies exploring single-ingredient interventions, findings are mixed. Several studies have supported the clinical utility of omega-3 PUFAs for cognitive enhancement and reduced dementia risk (5, 7); while, other similarly designed studies have contradicted such purported benefits (8). These seemingly incompatible reported findings served as a point of departure for a more systematic investigation. To this end, a Cochrane review (3) assimilated data from three interventional studies investigating the impact of omega-3 PUFA (EPA-DHA) supplementation on cognitive function in healthy older adults. The results refuted the purported benefits of omega-3 PUFAs on cognitive function following supplementation of 700mg/day EPA-DHA over 24-months (10), 400mg/day EPA-DHA over 40-months (11), and 1800mg/day or 400mg/day EPA-DHA six-months (12). In contrast, Witte and colleagues (13) assessed the impact of 26-week supplementation of 2200mg/day EPA-DHA on cognitive function in healthy older adults and observed measureable enhanced executive functions in the treatment group. Moreover, a double-blind placebo-controlled proof-of-concept trial found a differential beneficial effect of 2200mg/day omega-3 PUFA supplementation over 26 weeks on recall in an object-location-memory task but not for AVLT performance (14). However, the present study used a similar research design, omega-3 PUFA intervention dose and duration, and comparable cognitive assessment battery, but did not yield concordant results. The present study provided limited evidence for the positive and prophylactic impact of the multi-ingredient intervention (including omega 3 PUFA, vitamin D, resveratrol and whey protein), for maintaining neuronal health in later life.
The effects of vitamin D are also unclear from the previous literature, while some recent research has claimed a beneficial role of vitamin D in neuronal function (17); in contrast, other research reports no association between vitamin D status and cognitive function (18) or decline in cognitive performance over time (21). Here, mixed findings may be attributed to the fact that these studies did not use an interventional design and featured limited cognitive assessment batteries that may have lacked sensitivity for detecting subtle changes in cognitive function in healthy participants. The present study employed a prospective, longitudinal design with double-blind and placebo-controlled contrasts but reports negligible cognitive enhancement by the supplementation investigated.
Studies examining the impact of resveratrol on cognitive function remain in their infancy. Emerging animal and in vitro research suggests that dietary resveratrol may protect against cognitive decline in later life (22, 24). However, human clinical trials in this field are scarce. Moreover, evidence is lacking on the role of dairy constituents, such as whey protein, in cognition (28). Thus, the findings of the present study make an important contribution in this regard too; our results show that the combined omega 3 PUFA, vitamin D, resveratrol and whey protein supplementation did not yield benefits to cognitive function in this older adult sample.
A key challenge in the present study concerned retaining older adult participants in the longitudinal trial; seven participants dropped out before their intermediate assessment and a further seven, before their final assessment. Thus, it is possible that with a larger sample size, the associated increase in statistical power may have detected smaller effects. In addition, stringent recruitment procedures (namely, the pre-screening assessment of cognitive function via MMSE) may have favoured the selection of participants who were healthier than average i.e. the absence of a cognitive deficit. For instance, a recent investigation employing a broadly similar multi-ingredient nutrition supplement (omega-3 PUFAs, vitamin D, resveratrol and whey protein) reported improvements in cognitive function in older adults, albeit with a longer supplementation period and with participants with cognitive impairment ranging from mild to severe (43). Indeed, the majority of the present sample were university alumni. The recourse of educated participants is that the sample may have been unrepresentative of the wider older adult population. This may limit the generalizability of findings and raises issues from the standpoint of determining the efficacy of the intervention.
The study implemented a prospective, longitudinal design with double-blind and placebo-controlled contrasts to establish a causal effect of the intervention. In addition, standardised protocol was followed by trained and supervised researchers for data collection to reduce the potential impact of extraneous factors on cognitive performance. Participants were provided with a standard breakfast, and tested at the same time of day in the same room on both testing occasions. This allowed for the use of an extensive cognitive assessment battery comprising widely-used standardised measures with acceptable reliability and validity for use with the population under investigation. Finally, following previous research protocol, cognitive function was analysed by grouping crude individual test scores into a priori defined composite cognitive domains (12, 13, 30). The assimilation of cognitive measures in this way decreased variation associated with the individual tests, improved robustness of the outcomes and allowed for cross-comparison of findings with previous studies.
Although the present study reported no evidence elucidating the benefits of a combined omega-3 PUFA, vitamin D, resveratrol and whey protein intervention in this older adult sample, the results add to the large body of research in the field of nutrition, health and aging and extend the evidence base to an Irish context. From the perspective of identifying a suitable nutrition intervention to target age-related cognitive decline, current evidences are disappointing. Future researchers can build upon the current findings by conducting longer-term studies with larger more representative samples and incorporating diet and lifestyle measures, to more fully establish the prophylactic impact of the nutritional intervention on cognition.
In conclusion, the present study aimed to examine the impact of a targeted multi-ingredient nutrition supplement intervention, containing omega-3 PUFA, vitamin D, resveratrol and whey protein, on cognitive function. Overall, our findings suggest that the six-months of intervention had, with the exception of improved Stroop Color-Word performance, no beneficial impact on cognitive function in Irish community-dwelling older adults.


Acknowledgments: The authors’ responsibilities were as follows- JB, AM, BE, and GDV: designed the research; BE: acted as PI for the trial; CM and ASP: conducted data collection; JB & BR: planned and supervised the analysis; and CM and JB: wrote the manuscript, analysed the data and had primary responsibility for final content. All authors read and approved the final publication. A special word of thanks must be given to the psychology research assistants who assisted with data collection and recruitment; namely, Hannah Stynes, Susan Gibbons, Jane Maguire, Katie Grogan, Naoise MacGiollabhui, and Siobhan Blackwell. We would like to especially thank all the participants who generously contributed their time and efforts into realising this research.

Conflict of Interest: The research described herein was part-funded by Norwegian biotech company Smartfish®. Smartfish® provided the ready-to-drink juice formulations for the nutrition supplement intervention and placebo, but played no role in data collection, data analysis, or in the writing of this manuscript. None of the authors had any conflicts of interest with regard to the research described in this article.

Disclosure statementI: Catherine Moran had no conflicts of interest with regard to this research. Alessandro Scotto di Palumbo had no conflicts of interest with regard to this research. Jessica Bramham had no conflicts of interest with regard to this research. Aidan Moran had no conflicts of interest with regard to this research. Brendan Rooney had no conflicts of interest with regard to this research. Giuseppe De Vito had no conflicts of interest with regard to this research. Brendan Egan had no conflicts of interest with regard to this research.

Ethical standards:All study procedures were enacted in accordance with the ethical codes of conduct of the Psychological Society of Ireland and the guidelines of the Declaration of Helsinki (2008, 2013). The research protocol was granted ethical approval from the UCD Human Research Ethics-Sciences Board (reference: LS-13-28-Egan). Participants provided written informed consent prior to study enrolment. No animals were included in this research.



1. Johansson B. Memory and cognition in aging. In: Woods R, Clare L (eds) Handbook of the clinical psychology of aging, 2nd edn. John Wiley & Sons LTD, New York, 2008;pp 33-55.
2. Forte R, Boreham CA, Leite JC, De Vito G, Brennan L, Gibney ER, Pesce C. Enhancing cognitive functioning in the elderly: multicomponent vs resistance training. Clin Interv Aging 2013;8: 19-27. doi: 10.2147/CIA.S36514.
3. Sydenham E, Dangour AD, Lim WS. Omega 3 fatty acid for the prevention of cognitive decline and dementia. Cochrane Database of Syst Rev 2012;13(6): 1–42. doi: 10.1002/14651858.CD005379.pub3.
4. Gillette Guyonnet S, Abellan van Kan G, Andrieu S, Barberger Gateau P, Berr C, Bonnefoy M, Dartigues JF, de Groot L, Ferry M, Galan P, Hercberg S, Jeandel C, Morris MC, Nourhashemi F, Payette H, Poulain JP, Portet F, Roussel AM, Ritz P, Rolland Y, Vellas B. IANA task force on nutrition and cognitive decline with aging. J Nutr Health Aging 2007;11: 132– 152.
5. Barberger-Gateau P, Raffaitin C, Letenneur L, Berr C, Tzourio C, Dartigues JF, Alperovitch A. Dietary patterns and risk of dementia: the Three-City cohort study. Neurology 2007;69: 1921–1930. doi: 10.1212/01.wnl.0000278116.37320.52
6. Hooijmans CR, Pasker-de Jong PC, de Vries RB, Ritskes-Hoitinga M. The effects of long-term omega-3 fatty acid supplementation on cognition and Alzheimer’s pathology in animal models of Alzheimer’s disease: a systematic review and meta-analysis. J Alzheimers Dis 2012;28: 191–209. doi: 10.3233/JAD-2011-111217.
7. van Gelder BM, Tijhuis M, Kalmijn S, Kromhout D. Fish consumption, n-3 fatty acids, and subsequent 5-y cognitive decline in elderly men: the Zutphen Elderly Study. Am J Clin Nutr 2007;85: 1142–1147.
8. Engelhart MJ, Geerlings MI, Ruitenberg A, van Swieten JC, Hofman A, Witteman JC, Breteler MM. Diet and risk of dementia: does fat matter? The Rotterdam Study. Neurology 2001;59: 1915–1921. doi: 10.1212/01.WNL.0000038345.77753.46.
9. Kalmijn S, Feskens EJ, Launer LJ, Kromhout D. Polyunsaturated fatty acids, antioxidants, and cognitive function in very old men. Am J Epidemiol 1997;145: 33–41.
10. Dangour AD, Allen E, Elbourne D, Faey N, Fletcher AE, Hardy P, Holder GE, Knight R, Letley L, Richards M, Uauy R. Effect of 2-y n-3 long-chain polyunsaturated fatty acid supplementation on cognitive function in older people: a randomised, double-blind, controlled trial. Am J Clin Nutr 2010;91: 1725–1732. doi: 10.3945/ajcn.2009.29121.
11. Geleijnse J, Giltay E, & Kromhout D. Effects of n-3 fatty acids on cognitive decline: A randomised, double-blind, placebo-controlled decline in stable myocardial infarction patients. Alzheimers Dement 2012;8: 278–287. doi: 10.1016/j.jalz.2011.06.002.
12. van de Rest O, Geleijnse JM, Kok FJ, van Staveren WA, Dullemeijer C, Olderikkert MGM, Beekman AT, de Groot CP. Effect of fish oil on cognitive performance in older subjects; A randomised, controlled trial. Neurology 2008;71: 430–438. doi: 10.1212/01.wnl.0000324268.45138.86.
13. Witte AV, Kerti L, Hermannstädter HM, Fiebach JB, Schreiber SJ, Schuchardt JP, Hahn A, Flöel A. Long-chain omega-3 fatty acids improve brain function and structure in older adults. Cereb Cortex 2014;24: 3059-3068. doi: 10.1093/cercor/bht163.
14. Külzow N, Witte VA, Kerti L, Grittner U, Schuchardt JP, Hahn A, Flöel A. Impact of omega-3 fatty acid supplementation on memory functions in healthy older adults. J Alzheimers Dis 2016;51: 713-725. doi: 10.3233/JAD-150886
15. Buell JS, Dawson-Hughes B. Vitamin D and neurocognitive dysfunction: Preventing “D”ecline? Mol Aspects Med 2008;29: 415–422. doi: 10.1016/j.mam.2008.05.001.
16. Miller JW. Vitamin D and cognitive function in older adults. Are we concerned about vitamin D-mentia? Neurology 2010;74: 13–15. doi: 10.1212/WNL.0b013e3181c719a2.
17. Llewellyn DJ, Lang IA, Langa KM, Muniz-Terrera G, Phillips CL, Cherubini, A, Ferrucci L, Melzer D. Vitamin D and risk of cognitive decline in elderly persons. Arch Inter Med 2010;170: 1135–1141. doi: 10.1001/archinternmed.2010.173.
18. Slinin Y, Paudel ML, Taylor BC, Fink HA, Ishani A, Canales MT, Yaffe K, Barrett-Connor E, Orwoll ES, Shikany JM, Leblanc ES, Cauley JA, Ensrud KE, Osteoporotic Fractures in Men (MrOS) Study Research Group. 25-Hydroxyvitamin D levels and cognitive performance and decline in elderly men. Neurology 2010;74: 33–41. doi: 10.1212/WNL.0b013e3181c7197b.
19. Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12: 189–198. doi: 10.1016/0022-3956(75)90026-6.
20. Army Individual Test Battery. Manual of directions and scoring. War Department, Adjutant General’s Office, Washington DC, 1944.
21. Laughlin GA, Kritz-Silverstein D, Bergstrom J, Reas ET, Jassal SK, Barrett-Connor E, McEvoy LK. Vitamin D insufficiency and cognitive function trajectories in older adults: The Rancho Bernardo study. J Alzheimers Dis 2017;58: 871-883. doi: 10.3233/JAD-161295.
22. Anekonda TS. Resveratrol – A boon for treating Alzheimer’s disease? Brain Res Rev 2006;52: 316–326. doi: 10.1016/j.brainresrev.2006.04.004.
23. Bedalov A, Simon JA. Neuroscience. NAD to the rescue. Science 2004;305: 954-955.
24. Parker JA, Arango M, Abderrahmane S, Lambert E, Tourette C, Catoire H, Neri C. Resveratrol rescues mutant polyglutamine cytotoxicity in nematode and mammalian neurons. Nat Genet 2005;37: 349–350. doi: 10.1038/ng1534.
25. Walle T. Bioavailability of resveratrol. Ann N Y Acad Sci 2011;1215: 9–15. doi: 10.1111/j.1749-6632.2010.05842.x.
26. Bauer JM, Verlaan S, Bautmans I, Brandt K, Donini LM, Maggio M, McMurdo ME, Mets T, Seal C, Wijers SL, Ceda GP, De Vito G, Donders G, Drey M, Greig C, Holmbäck U, Narici M, McPhee J, Poggiogalle E, Power D, Scafoglieri A, Schultz R, Sieber CC, Cederholm T . Effects of a vitamin D and leucine-enriched whey protein nutritional supplement on measures of sarcopenia in older adults, the PROVIDE Study: A randomized, double-blind, placebo-controlled trial. J Am Med Dir Assoc 2015;16(9): 740-747. doi: 10.1016/j.jamda.2015.05.021.
27. Weaver CM. Role of dairy beverages in the diet. Physiol Behav 100: 2010;63–66. doi: 10.1016/j.physbeh.2010.01.020.
28. Camfield DA, Owen L, Scholey AB, Pipingas A, Stough C. Dairy constituents and neurocognitive health in aging. Br J Nutr 2011;106: 159–174. doi: 10.1017/S0007114511000158.
29. Greig CA, Young A, Skelton DA, Pippet E, Butler FM, Mahmud SM. Exercise studies with elderly volunteers. Age Ageing 1994;23: 185–189. doi: 10.1093/ageing/23.3.185.
30. Beauchet O, Annweiler C, Assal F, Bridenbaugh S, Herrmann FR, Kressig RW, Allali G. Imagined Timed Up & Go test: A new tool to assess higher-level gait and balance disorders in older adults? J Neurol Sci 2010;294: 103-106. doi: 10.1016/j.jns.2010.03.021.
31. Broadbent DE, Cooper PF, FitzGerald P, Parkes KR. The Cognitive Failures Questionnaire (CFQ) and its correlates. Brit J Clin Psychol 1982;21: 1-16. doi: 10.1111/j.2044-8260.1982.tb01421.x.
32. Rey A. The clinical examination in psychology. Universitaires de France, Paris, 1964.
33. Stroop J. Studies of interference in serial verbal reactions. J Exp Psychol 1935;18: 643–662. doi: 10.1037/h0054651.
34. Trenerry M, Crosson B, DeBoe J, Leber W. Stroop neuropsychological screening test manual. Psychological Assessment Resources (PAR), Adessa, Florida, 1989.
35. Benton AL, Hamsher KD. Multilingual aphasia examination. AJA Associates, Iowa City, Iowa, 1989.
36. Wechsler D. Wechsler adult intelligence scale, 3rd edn. Harcourt Assessment, San Antonio, Texas, 1997
37. Jennerod M. Mental imagery in the motor context. Neuropsychologia 1995;33: 1419-1432. doi: 10.1016/0028-3932(95)00073-C.
38. Podsiadlo D, Richardson S. The timed “up and go”: a test of basic functional mobility for frail elderly persons. J Am Geriatr Soc 1991;39: 142-148. doi: 10.1111/j.1532-5415.1991.tb01616.x.
39. Beauchet O, Launay CP, Sejdic E, Allali G, Annweiler C. Motor imagery of gait: a new way to detect mild cognitive impairment? J Neuroeng Rehabil 2014;11(66): 1-7. doi: 10.1186/1743-0003-11-66.
40. Vogel A, Stokholm J, Gade A, Andersen BB, Hejl AM, Waldemar G. Awareness of deficits in mild cognitive impairment and Alzheimer’s disease: Do MCI patients have impaired insight? Dement Geriatr Cogn Disord 2004;17(3): 181-187. doi: 10.1159/000076354.
41. IBM SPSS Statistics. IBM SPSS statistics 20.0 for Windows/Apple Mac. SPSS Inc, Chicago, 2011.
42. Norman G. Likert scales, levels of measurement and the “laws of statistics”. Adv Health Sci Educ Theory Pract 2010;15: 625–632. doi: 10.1007/s10459-010-9222-y.
43. Famenini S, Rigali EA, Olivera-Perez HM, Dang J, Chang MT, Halder R, Rao RV, Pellegrini M, Porter V, Bredesen D, Fiala M. Increased intermediate M1-M2 macrophage polarization and improved cognition in mild cognitive impairment patients on omega-3 supplementation. FASEB J 2017;31: 148-160. doi: 10.1096/fj.201600677RR.



G. La Fata1, N. van Vliet2, S. Barnhoorn2, R.M.C. Brandt2, S. Etheve1, E. Chenal1, C. Grunenwald1, N. Seifert1, P. Weber1, J.H.J. Hoeijmakers2,3, M.H. Mohajeri1,#, W. P. Vermeij2,#


1. DSM Nutritional Products Ltd., P.O. Box 2676, CH-4002 Basel, Switzerland; 2. Department of Molecular Genetics, Erasmus University Medical Center Rotterdam, P.O. Box 2040, 3000 CA, Rotterdam, The Netherlands; 3. CECAD Forschungszentrum, Universität zu Köln, Köln, Germany. # Equal contribution 

Corresponding Author: M. Hasan Mohajeri, DSM Nutritional Products Ltd., P.O. Box 2676, CH-4002 Basel, Switzerland,; Second corresponding Author: Wilbert P. Vermeij, Department of Molecular Genetics, Erasmus University Medical Center Rotterdam, P.O. Box 2040, 3000 CA, Rotterdam, The Netherlands,

J Prev Alz Dis 2016 inpress
Published online August 8, 2017,



Background: Aging is a highly complex biological process driven by multiple factors. Its progression can partially be influenced by nutritional interventions. Vitamin E is a lipid-soluble anti-oxidant that is investigated as nutritional supplement for its ability to prevent or delay the onset of specific aging pathologies, including neurodegenerative disorders.
Purpose: We aimed here to investigate the effect of vitamin E during aging progression in a well characterized mouse model for premature aging.
Method: Xpg-/- animals received diets with low (~2.5 mg/kg feed), medium (75 mg/kg feed) or high (375 mg/kg feed) vitamin E concentration and their phenotype was monitored during aging progression. Vitamin E content was analyzed in the feed, for stability reasons, and in mouse plasma, brain, and liver, for effectiveness of the treatment. Subsequent age-related changes were monitored for improvement by increased vitamin E or worsening by depletion in both liver and nervous system, organs sensitive to oxidative stress.
Results: Mice supplemented with high levels of vitamin E showed a delayed onset of age-related body weight decline and appearance of tremors when compared to mice with a low dietary vitamin E intake. DNA damage resulting in liver abnormalities such as changes in polyploidy, was considerably prevented by elevated amounts of vitamin E. Additionally, immunohistochemical analyses revealed that high intake of vitamin E, when compared with low and medium levels of vitamin E in the diet, reduces the number of p53-positive cells throughout the brain, indicative of a lower number of cells dying due to DNA damage accumulated over time.
Conclusions: Our data underline a neuroprotective role of vitamin E in the premature aging animal model used in this study, likely via a reduction of oxidative stress, and implies the importance of improved nutrition to sustain health.

Key words: Vitamin E, neurodegeneration, aging, DNA damage repair, anti-aging interventions.



Aging is a highly complex biological process whose progression is generally associated with time-dependent accumulation of cellular damage (1, 2). For instance, the integrity of DNA is continually challenged by exogenous and endogenous threats including reactive oxygen species (ROS), leading to time-dependent accumulation of DNA lesions, which interfere with vital DNA metabolism and may cause cellular dysfunction and cell death (3, 4). At high concentrations, ROS are detrimental and responsible for compromising normal cellular functioning (5). Young and healthy cells are equipped with complex anti-oxidant defense systems to prevent the high reactivity of ROS, and various DNA repair mechanisms to repair (oxidative) DNA lesions (4, 5).
One such DNA repair system is the nucleotide excision repair (NER) pathway that is primarily responsible for clearing helix-distorting DNA lesions (6). This pathway consists of two branches using a distinct manner of damage recognition. One scans the whole genome for lesions and is named global genome (GG) NER while the other is triggered by transcription stalled by DNA damage and therefore called transcription-coupled NER. In total, NER employs over 30 proteins, including the endonuclease XPG (also known as ERCC5). In both sub-branches of NER, XPG cuts the DNA 5-6 nucleotides downstream of the damage, as part of a dual incision mechanism for excision of the damage (6, 7). Additionally, XPG has also been implicated in other processes such as homologous recombination repair of double stranded DNA breaks and removal of small oxidative DNA lesions by base excision repair (8, 9). Mutations that affect activity or expression levels of the endonuclease XPG are rare and cause a spectrum of severe symptoms, including striking hypersensitivity to sun exposure, early cessation of growth, accelerated aging features (including progressive neurodegeneration and body weight decline) and reduced life expectancy (10) a constellation of features characteristic of the rare transcription-coupled NER condition Cockayne syndrome (CS). Some XPG mutations strongly predispose patients to cancer and pigmentation abnormalities in sun-exposed parts of the skin, typical of the rare (GG-) NER disorder xeroderma pigmentosum (XP), or a combination of XP/CS.
Recently, a new DNA repair-deficient mutant mouse (Xpg-/-) was generated to mimic the genetic defects in Cockayne syndrome-type XP-G patients. This model is characterized by shortened lifespan, of about 18 weeks, and accelerated onset of multiple progressive aging features such as prominent neurodegeneration, including loss of hearing, vision and motor performance, cognitive decline, early development of tremors, imbalance and paresis, as well as cachexia, age-related abnormalities in multiple tissues and organs including liver, kidney, skeleton, intestine, skin, retina, muscles, vascular system and heart and overall frailty, strikingly resembling Cockayne syndrome in man (11, 12) (and unpublished results).
Alterations in macronutrients can have a great impact on health and aging of mice and other organisms (13). Xpg-/- mice are very sensitive to nutritional changes and respond remarkably well to dietary interventions (14). Indeed, a reduction of feed intake by 30%, known as diet restriction , that in normally aging wild type mice would yield in about 30% lifespan extension, drastically extended the lifespan of these animal models by 80% and delayed the onset of many age-related neurological abnormalities, such as tremors, imbalance and paresis (14).
Vitamin E includes eight structurally-related lipid-soluble compounds with potent anti-oxidant properties consisting of four tocopherols and four tocotrienols: α (alpha), β (beta), γ (gamma) and δ (delta) (15). α-Tocopherol is the most abundant and bioavailable form of vitamin E in human and rodents (16, 17) and serves as a strong ROS scavenger that protects polyunsaturated fatty acids in the cellular membranes and in lipoproteins from peroxidation (18).
Since vitamin E is known as potent anti-oxidant, we hypothesize that supplementing Xpg-/- mice with vitamin E might reduce some of the aging features driven by (oxidative) DNA damage accumulation. Although a number of papers showed beneficial effects associated with vitamin E supplementation, few reports described inconsistent or negative results (19-21) and references within]. The difficulty in performing precise and uniform human studies resides in several important variables such as: age and health status of the individuals included in the study, nutritional status and diet, form of vitamin E administered, but also quantity and time of supplementation (19). Moreover, other important parameters to be considered are the genetic variations that may alter absorption, availability, metabolism, excretion and therefore net activity of these compounds. Working with animal models may reduce the variability associated with all these parameters and therefore in vivo studies represent excellent tools to unravel the function of specific nutrients during aging progression.


Materials and Methods

Mouse model

The generation, genotyping and characterization of Xpg-/- mice has been previously described (11). Xpg-/- mice were obtained by crossing Xpg+/- mice with a pure C57BL6J and pure FVB backgrounds to yield Xpg-/- pups with a F1 C57BL6J/FVB hybrid background. Typical unfavorable characteristics, like blindness in an FVB background or deafness in a C57BL6J background, do not occur in this hybrid background (11).

Housing conditions

Since Xpg-/- mice are smaller than their wildtype littermates, feed was administered within the cages and water bottles with long nozzles were used. Animals were maintained in a controlled environment (20–22°C, 12h light : 12h dark cycle) and were housed in individual ventilated cages under Specific Pathogen-Free conditions. All animals were bred on AIN93G synthetic pellets (Research Diet Services B.V., Wijk bij Duurstede, The Netherlands; gross energy content 4.9 kcal/g dry mass, digestible energy 3.97 kcal/g; with choline bitartrate replaced for choline chloride to avoid potential formation of bladder and kidney stones).

Study description

Immediately after birth, litters together with their mothers were distributed into three groups receiving three diets taking into account: day of birth, litter size, and genetic background of the mother. All diets were relabeled before entering the mouse facility and given in a blinded manner. After weaning, at 4 weeks of age, all mice were kept on the designated diet and were individually housed to accurately measure feed intake. The diets were: a) AIN93G synthetic pellets without the addition of any vitamin E (hereafter referred to as “low”, considering this diet still contains traces of vitamin E from the Casein protein source); b) AIN93G pellets containing medium levels (75 mg/kg) of vitamin E (all-rac-α-tocopheryl acetate; hereafter referred to as “medium”, which is normally present in this diet, and is slightly higher than the recommended dosage for mice); c) AIN93G pellets with addition of extra vitamin E to a final concentration of 375 mg/kg (hereafter referred to as “high”).  Mice were analyzed at 8 and 12 weeks of age (Figure 1a).

a) Schematic view of the study design. The three diets used in the study are indicated by the horizontal elongated rectangles: white (low vitamin E diet: traces), grey (medium vitamin E diet: 75 mg/kg), black (high vitamin E diet: 375 mg/kg). Lower black line represents timeline with black dots representing key events during the study. b-d) α-Tocopherol measurements in the three diets: low, medium, high respectively. “Start” and “End” indicate the beginning and the final stage of the supplementation study. Error bars represent standard error of the mean (s.e.m.).

Phenotype scoring: body weight and tremors

Mice were clinically examined on a daily basis by the animal caretakers. Moreover, they were weighed, visually inspected and phenotypically scored for age related symptoms in a blinded manner by two experienced research technicians. The onset of body weight loss was counted as the first week a decline in body weight was noted after their maximal body weight was reached. Whole body tremor was scored if mice showed trembling for a combined total of at least 10 seconds when put on a flat surface for 20 seconds.

Micronutrients measurements

Chemicals and reagents

Retinol (vitamin A), α-tocopherol (vitamin E), thiamin hydrochloride (vitamin B1), riboflavin (vitamin B2) and pyridoxal-5’-phosphate (vitamin B6) were purchased from Sigma-Aldrich. The corresponding deuterated internal standard, retinol-d5, thiamin-d3, riboflavine 13C4 15N2 and pyridoxal-d3 5-phosphate were purchased from medical isotopes and vitamin A-d5 acetate from Alsachim. Other chemicals used were analytical grade and obtained from either Sigma-Aldrich or Merck Millipore. Water used was passed through a Milli-Q water purification system (Millipore).


Separation and quantification was performed on an Agilent 1290 ultra-high performance liquid chromatography (UHPLC) coupled with an API 4000 mass spectrophotometer (MS) from AB Sciex using multiple reaction monitoring (MRM) transitions. A turbo ion spray source operating in positive mode was used for water soluble vitamins determination and a photospray ionization source operating in positive mode was used for fat soluble vitamins determination.
For B-vitamin analysis, the analytical column was an Ascentis Express C8 from Supelco. Samples were eluted using a gradient from 100% of acidified water (with acetic acid) to 100% of acetonitrile, with help of heptafluorobutiric acid used as ion pairing reagent.
For vitamin A and vitamin E determination, the analytical column was a Halo C18 from Advanced material technologies. Samples were eluted using a gradient from 8% of acidified water (with formic acid) to 100% of an acidified methanol/acetonitrile solution.

Sample preparation

Water soluble vitamins in plasma, brain and liver: in order to remove the protein and extract the B-vitamins, an aliquot of plasma sample or a part of the tissues (brain or liver) was combined with a solution of trichloroacetic acid (50g/L) containing the internal standards. After centrifugation (and filtration for tissue samples) the supernatant was injected on the UHPLC-MS system.
Vitamin A and vitamin E in plasma: after addition of internal standards to an aliquot of plasma, proteins were removed with the addition of a mixture of tetrahydrofuran, acetonitrile and methanol. After filtration and centrifugation, the supernatant was injected on the UHPLC-MS system.
Vitamin A and vitamin E in liver: internal standards and vitamin C (as anti-oxidant) were added to the tissue. A saponification was conducted with a methanol/potassium hydroxide solution, followed by a liquid-liquid extraction with hexane and a concentration step, the resulting supernatant was injected on the UHPLC-MS system.
Vitamin A and vitamin E in brain: internal standards and water/ethanol solution were added to the tissue samples. After a liquid-liquid extraction with hexane and concentration step, the resulting supernatant was injected on the UHPLC-MS system.


Identification and quantification were performed applying external calibrations (using dedicated internal standards). Moreover, quality control samples were prepared with the similar procedure as the unknown samples and analyzed at each run. These methods enable the measurement of water and fat soluble vitamins with good accuracy (85-115%), inter and intra-day precision (<15%) in plasma, liver and brain. In addition, lower limits of quantitation (LOQ) were determined for tocopherol (250 ng/mL in plasma and 5.00 ng/mg in tissue), retinol (25.0 ng/mL in plasma and 0.50 ng/mg in tissue), thiamine (5.00 ng/mL in plasma and 0.050 ng/mg in tissue), riboflavin (0.5 ng/mL in plasma and 0.020 ng/mg in tissue) and vitamin B6 (0.5 ng/mL in plasma and 0.020 ng/mg in tissue).

Vitamin E measurements in the feed

Vitamin E in feed was analyzed as reported in (22). Briefly after saponification with with ethanolic potassium hydroxide, the unsaponifiable residue of the sample is extracted with cyclohexane/diethyl ether. Quantification is carried out with normal phase HPLC and fluorescence detection using α-tocopherol as external standard. With this analytical method, the sum of free as well as esterified α-tocopherol (α-tocopheryl acetate) is determined.

Markers of oxidative stress

Redox state of thiol groups

Two independent research groups previously developed an assay for monitoring thiol redox states (23, 24). Both assays used chemical modifiers that can specifically bind covalently with, and thereby inactivate, free thiol groups. Reducing agents cannot remove these modifications. After subsequent reduction of oxidized thiol group, that were engaged in disulfide bonds, these could be labeled with a second compound, to monitor thiol-based redox changes using mass spectrometry or histology (23, 24). Both protocols were here combined into an immunoblotting-based assay: To quantify the SS/SH redox ratio ~20 mg of liver tissue was lysed and sonicated in 400 µl “reduced thiol labeling solution”, containing 50 mM Tris-Cl pH 7.5, 1% SDS, 1 mM EDTA, 10 mM N-Ethylmaleimide (NEM; Sigma-Aldrich), 20 µM Alexa 800 maleimide (Thermo Fisher Scientific), and 1 protease inhibitor cocktail tablet (Roche) per ml dH2O for 50x conc (25). After 5 minutes heating at 70°C and 30 minutes incubation at room temperature the labeling was quenched by adding fresh NEM to a final concentration of 100 mM. 100 µl homogenate was precipitated in 1 ml cold Acetone:MeOH:EtOH (2:1:1) and re-suspended in 50 mM Tris-Cl pH 7.5, 1% SDS, 10 mM TCEP (Sigma-Aldrich) by vigorously pipetting up and down. Samples were heated for 5 minutes at 70°C, incubated for 30 minutes at room temperature and precipitated again with 1 ml cold Acetone:MeOH:EtOH (2:1:1). Pellets were re-dissolved in 200 µl “oxidized thiol labeling solution”, containing 50 mM Tris-Cl pH 7.5, 1% SDS, 10 mM NEM, and 20 µM Alexa 680 maleimide (Thermo Fisher Scientific), to label the thiols groups previously engaged in disulfide bonds. After 30 minutes incubation at room temperature and another round of precipitation, samples were dissolved in SDS-sample buffer and run on PAGE. Gel was fixed in 1% Orto-phosphoric acid, 50% EtOH and scanned using the Odyssey (Li-Cor Biosciences, USA). The ratio between the signal intensity per lane was quantified using the instruments software.

Protein carbonylation

For determination of the protein carbonyl content ~20 mg of liver tissue was lysed in 400 µl SDS-sample buffer. 50 µg homogenate was transferred to a new tube, supplemented with SDS to a total concentration of 12%, and incubated for 30 minutes at room temperature with 2 volumes of 20 mM 2,4-Dinitrophenylhydrazine (DNPH; Sigma-Aldrich) in 10% Trifluoroacetic acid (TCA) or 2 volumes 10% TCA only as negative control. The reaction was neutralized with 1.7 starting volume of 2M Tris base containing 30% glycerol and the protein samples were transferred by dotblot to PVDF membrane (Millipore). Carbonylation was visualized with anti-DNPH (Sigma-Aldrich; D8406; 1:2,000) in 5% BSA in PBS-Tween 0.05% and rat anti-mouse IgE-HRP (SouthernBiotech; 1130-05; 1:5,000), using ECL+ (PerkinElmer) on the ImageQuant LAS4000 mini (GE Healthcare Life Sciences). Signal intensities were quantified in ImageQuant TL software and corrected for total protein by Ponceau S. All samples were analyzed at least in duplo.

Histological analysis of nuclear DNA content

Fresh frozen liver tissue was embedded in TissueTek and sliced in 10 µm thick cryosections. Sections were shortly dried, post-fixed for 10 minutes in 4% paraformaldehyde, and stained with hematoxylin. Images were generated using the NanoZoomer Digital slide scanner (Hamamatsu Photonics, Japan) and the nuclei size was quantified using NDP view software in a blinded manner. A similar number of images was quantified for each group.

Oil Red O staining

Tissues were snap-frozen in liquid nitrogen, embedded in TissueTek and sliced in 10 µm thick cryosections and mounted on Superfrost Plus slides. Oil red O solution was applied to liver sections for 5 min. Slides were washed twice in water, 15 min each wash, and counterstained with hematoxylin. Oil red O images were generated using the NanoZoomer Digital slide scanner (Hamamatsu Photonics, Japan) and the Oil Red O intensity was quantified using Fiji software.

TUNEL staining

To quantify apoptotic cells in the retina, eyes were fixed overnight in 10% phosphate-buffered formalin, paraffin-embedded, sectioned at 5 µm, and mounted on Superfrost Plus slides. Paraffin sections were employed for TdT-mediated dUTP Nick-End Labeling (TUNEL) assay using a commercial kit (Apoptag Plus Peroxidase in situ apoptosis detection kit, Millipore). Sections were deparaffinized and incubated as described by the manufacturer.

Histological examination of neurodegeneration

Mice were anaesthetized with pentobarbital and transcardially perfused with 4% paraformaldehyde. Brains were carefully dissected out, post-fixed for 1.5 h in 4% paraformaldehyde, cryoprotected, embedded in 12% gelatin, rapidly frozen, and sectioned at 40 µm using a freezing microtome or stored at -80°C until use. Brain tissues of mice from different dietary regimes were embedded together, in one gelatin block per time point, to avoid fluctuations in antibody affinity. To quantitatively assess protein regulation, all immunohistochemical procedures were performed simultaneously for each antibody.
The extent of astrogliosis was assessed using rabbit anti-GFAP (DAKO; Z0334; 1:15,000) as primary antibody with a biotinylated secondary antibody from Vector Laboratories (BA1000; 1:200). All sections were processed free floating using the ABC method (ABC, Vector Laboratories, USA) with diaminobenzidine (0.05%; Sigma-Aldrich) as the chromogen. Immunoperoxidase-stained sections were imaged using the NanoZoomer Digital slide scanner with the NDP view software (Hamamatsu Photonics, Japan). Mean intensities were quantified using Fiji software.
The extent of neuronal genomic stress was assessed using rabbit anti-p53 (Leica; NCL-p53-CM5p; 1:1,000) with Cy3-conjugated goat-anti-rabbit (Jackson ImmunoResearch; 111-165-144; 1:200) and DAPI for cell nuclei. Immunofluorescence sections were analyzed using a Zeiss LSM700 confocal microscope. P53 quantification was performed using Fiji software.


Statistical differences were calculated using a one-way ANOVA including Bonferroni’s multiple comparison test or Dunnett t (2-sided) using IBM SPSS Statistics 21. For the vitamin E quantification in the feed (two groups), and the vitamin measurements in liver and brain (too low n) statistics was assessed with a t-test. Plasma vitamin levels were assessed with a one-way ANOVA and Bonferroni’s multiple comparison test. Onset of tremors and body weight decline were statistically analyzed with survival curve analysis using the product limit method of Kaplan and Meier with Log-rank (Mantel-Cox) test in GraphPad Prism. Significance is indicated in the tables and figures by (*) p values < 0.05; (**) p values <0.01; (***) p values <0.001.



Concentration of vitamin E and other micronutrients

We first measured the concentration of α-Tocopherol in the three diets at the beginning and at the end of the study (Figure 1b-d). As expected, the measured values were in agreement with the theoretical concentrations (~2.5, 75, and 375 mg/kg feed respectively) and we did not observe any change in vitamin E stability over time (Figure 1b-d, compare “Start” with “End” stage). Secondly, we examined the concentration of vitamin E, and a few additional other micronutrients that potentially could affect brain development, in plasma, liver, and brain of the experimental animals at 8 and 12 weeks. Increasing values of vitamin E in the feed, corresponded in all three tissues to significant increasing values of vitamin E specifically (Table 1 and Table 2).

Figure 1. Study design and vitamin E content in the three diets

Figure 1. Study design and vitamin E content in the three diets

a) Schematic view of the study design. The three diets used in the study are indicated by the horizontal elongated rectangles: white (low vitamin E diet: traces), grey (medium vitamin E diet: 75 mg/kg), black (high vitamin E diet: 375 mg/kg). Lower black line represents timeline with black dots representing key events during the study. b-d) α-Tocopherol measurements in the three diets: low, medium, high respectively. “Start” and “End” indicate the beginning and the final stage of the supplementation study. Error bars represent standard error of the mean (s.e.m.).


Table 1. Vitamin E levels in Xpg-/- mice at 8 and 12 weeks

Table 1. Vitamin E levels in Xpg-/- mice at 8 and 12 weeks

Amount of α-Tocopherol in the plasma (nmol/L, n=7), liver and brain (nmol/g, n=2 for all and n= 3 for high vitamin E diet at 12 weeks) of 8 and 12 weeks old Xpg-/- mice.  Errors represent s.e.m. values. a = compares low with medium vitamin E group; b = medium vs high;  c = high vs low; ns: not significant.


Table 2. Micronutrient levels in Xpg-/- mice at 8 and 12 weeks

Table 2. Micronutrient levels in Xpg-/- mice at 8 and 12 weeks

Amount of retinol (vitamin A), thiamin (vitamin B1), riboflavin (vitamin B2) and pyridoxal 5’phosphate (vitamin B6) in the plasma (nmol/L), liver and brain (nmol/g) of 8 and 12 weeks old Xpg-/- mice. Values represent averages of n>2 biological replicates: plasma n=7, liver and brain n=2 and n=3 for high vitamin E diet at 12 weeks. Errors represent s.e.m. values.


Feed intake, body weight and onset of tremors

No clear differences were observed in average feed intake by the lack or increase of vitamin E (Figure 2a and b; females and males respectively), excluding indirect effects of dietary restriction for which Xpg-/- mice are very sensitive (14). Concomitantly, no obvious differences in average body weight of Xpg-/- mice fed with diets containing different dosages of vitamin E were measured (Figure 2c and d; females and males respectively). Only a minor difference in body weight at 4 weeks of age could be observed, likely due to the random distribution of litters at birth over the different diets (Figure 2c). Subsequently, both females and males gained in body weight until approximately 8 weeks of age. After reaching their maximal body weight they all gradually declined with age as consequence of the aging-associated deterioration, which includes cachexia, like in Cockayne syndrome patients. The onset of body weight decline was significantly delayed in a vitamin E dose-dependent manner (Figure 2e). The animals receiving the highest amount of vitamin E in their feed showed a median delay of 2 weeks in onset of body weight decline compared to the animals with the lowest amount of vitamin E given (Figure 2e; p=0.0268).

Figure 2. Vitamin E prevents age-related decline in body weight and onset of tremors in Xpg-/- mice

Figure 2. Vitamin E prevents age-related decline in body weight and onset of tremors in Xpg-/- mice

a-b) Feed intake expressed in grams (g)/day measured weekly from 4 to 12 weeks (wks) of age in Xpg-/- females (a, n=4) and Xpg-/- males (b, n=6). c-d) Body weight curves of Xpg-/- female (c, n=4) and male (d, n=6) mice measured weekly. For all graphs (a-d) the “n” indicates the number of animals (for each diet group) available until the first time point (8 weeks of age, Figure 1a). e) Onset of body weight (BW) decline measured in the Xpg-/- mice designated for time point 2 (12 weeks of age). Per animal the first week of BW decline was expressed in percentage (%) of animals presenting a reduced body weight at a given time (in weeks). n=6-8 animals per group *p=0.0268 (between low and high vitamin E using the Log-rank Mantel-Cox test). f) Onset of tremors (neurological abnormality) with age in the Xpg-/- mice fed with different amounts of vitamin E in their feed given as % of animals with tremors. n=7-8 animals per group *p=0.0455 (between low and high vitamin E using the Log-rank Mantel-Cox test). Importantly, animals were scored in a fully blind manner (see Material & Methods section). For all graphs in Figure 2, error bars indicate s.e.m.


Appearance of tremors, a neurodegenerative aging feature occasionally observed in old animals, was also monitored during the visual examination of the Xpg-/- mice. About 40% of the mice fed on a diet containing low vitamin E developed tremors at an age of 11 weeks, while no signs of tremors were present yet in the mice fed on high content of vitamin E (Figure 2f; p=0.0455). However, at 12 weeks of age all mice showed signs of tremors, similarly to what was previously observed for these mice (11). This minor observational difference prompted us to further investigate aging features in more detail.

Oxidative status

Since vitamin E is known as a potent anti-oxidant, we verified whether increasing amounts of vitamin E were associated with reduced oxidative stress levels, using two distinct assays.  Preliminary evidence indicated that the overall intracellular redox state (SS/SH ratio) was significantly reduced in liver homogenates with medium and high concentrations of vitamin E in the feed (Figure 3a; p=0.000064 and p=000004 relatively compared to low vitamin E). This dose-dependent shift was, however, not observed for the specific detection of protein carbonyl groups (Figure 3b). Due to the limited amount of brain tissue available for direct measurements, we could not measure the oxidation status of brains and the analysis was therefore limited to only liver tissue.


Figure 3. High vitamin E reduces polyploidization in the liver of Xpg-/- mice

Figure 3. High vitamin E reduces polyploidization in the liver of Xpg-/- mice

a) Immuno-blot analysis of the overall liver redox state. The ratio between oxidized (SS) and reduced (SH) thiol groups was determined for low (white bar) and high (black bar) vitamin E relative to the level of normal (gray bar) vitamin E. b) Dot-blot analysis of the protein carbonyl content in liver samples of low (white bar) and high (black bar) amounts of vitamin E relative to normal (gray bar) vitamin E. For a) and b) n>2 liver pieces of 2-3 animals/group.  c-d) Representative images (c) and average nucleus size of hepatocytes (d, gray lines) in the liver of 12 weeks old Xpg-/- mice fed with low (triangle down), medium (circle), or high (triangle up) vitamin E in the diet. Equal numbers of images were quantified from n>3 liver slices of 2-3 animals/group. Each point corresponds to one nucleus. Scale bar = 25µm. Error bars represent s.e.m. Statistical significance is measured by one-way ANOVA (***p<0.001).


Cellular damage in liver

Considering the delayed body weight decline and onset of tremors associated with a diet containing high amount of vitamin E (Figure 2), specific markers of aging were studied more in detail. We first focused on the liver, an organ not only important for vitamin E metabolism, but also very sensitive to oxidative stress (26). With reduced levels of vitamin E, more polyploid hepatocyte nuclei were observed (Figure 3c). Quantification of similar areas of liver indeed yielded in an increased average nuclei size and thereby less total hepatocytes per section. In turn, supplementation with high vitamin E reduced the hepatocyte nuclei size, compared to both diets with low and medium vitamin E levels (Figure 3c and d; p=9.364E-8 and p=0.000043 respectively). We also quantified the percentage of cells containing an increased nuclear size double the normal value (set as bigger than 80µm2). This was reduced in a vitamin E dose dependent manner, 18.5% (low), 15.8% (medium), to 9.6% (high), concomitantly with an increased number of hepatocytes present in a similar area (356, 429, and 499 respectively). Additionally, fat accumulation in the liver did not result in any obvious changes by dietary alterations of vitamin E (Supplementary Figure).

Cellular damage in nervous system

Since Xpg-/- animals display numerous features of nervous system aging that primarily can be halted by nutritional changes like diet restriction (14), we next investigated the effects of vitamin E changes on retina and brain. TUNEL staining confirmed cell death in the outer nuclear layer of the retina at 12 weeks of age, but no improvement by vitamin E supplementation could be detected (Figure 4a and b). Changing vitamin E content in the diet also did not result in macroscopical changes in brain (data not shown), or immunohistological changes in the glial fibrillary acidic protein (GFAP) expression, a marker for astrocytosis  (Figure 4c and d).

Figure 4. Vitamin E does not affect retina degeneration and astrogliosis in Xpg-/- mice

Figure 4. Vitamin E does not affect retina degeneration and astrogliosis in Xpg-/- mice

a) Quantification of TUNEL-positive cells in the outer nuclear layer of retinal sections of 12 weeks old Xpg-/-  mice on low (white bar), medium (gray bar), and high (black bar) vitamin E (n=5 animals per group). b) Representative image of TUNEL staining in the eye of a Xpg-/- mouse on low vitamin E diet. Scale bar = 500µm. Dying cells (TUNEL +) are indicated by black arrowheads in the higher magnification on the right. Scale bar = 50µm. c) Quantification of the relative intensity of consecutive sagittal brain sections immunoperoxidase-stained for GFAP. Brains from animals on low (white bar), medium (gray bar), and high (black bar) vitamin E diets were analyzed. The amount of signal per slice for the normal vitamin E diet was set on 1 (n=3 animals per group). d) Representative picture of a sagittal brain section of a 12 weeks old Xpg-/- mouse on low vitamin E immunoperoxidase-stained for GFAP. Scale bar = 1mm. For all graphs in Figure 4, error bars indicate s.e.m.


As marker for genotoxic stress in the nervous system, we subsequently studied the expression of transcription factor p53. Immunohistochemical analysis showed a significant reduced percentage of p53-positive cells in the brain of animals fed with high vitamin E content (Figure 5a). The reduced cellular degeneration in the brain was already observed in 8 weeks old Xpg-/- mice (Figure 5b, p=0.025 low vs. high and p=0.023 medium vs. high) and persisted in older animals (Figure 5c, p=0.053 low vs. high and p=0.042 medium vs. high). Of note, no differences were observed when brains from animals fed on low vitamin E diet were compared to those fed on a diet containing medium vitamin E levels (Figure 5b and c).


Figure 5. Vitamin E reduces the number of p53 positive cells in the brains of Xpg-/- mice

Figure 5. Vitamin E reduces the number of p53 positive cells in the brains of Xpg-/- mice

a) Representative images showing the genotoxic stress biomarker p53 in brain slices of Xpg-/- mice animals at 12 weeks of age fed on diets containing low, medium and high vitamin E. Upper panels show the DAPI staining (blue = cellular nuclei). Lower panels show the p53-positive cells (red staining). Enlargements show zoom areas with maximized contrast to visualize p53-positive cells (white arrows). Scale bar = 1mm. b-c) Quantification of the % of p53-positive cells in the brain of 8 weeks (b) and 12 weeks (c) old Xpg-/- mice fed on a diet with low, medium and high vitamin E content. For all analyses n=4. Error bars represent the s.e.m. Statistical significance is measured by one-way ANOVA (*p<0.05).  The number of p53-positive cells in Xpg-/- mice fed with medium vitamin E diet was set as 100%.



The results of this study show for the first time, that modifying the concentration of a single micronutrient (vitamin E) in the diet of the Xpg-/- premature aging murine model, was sufficient to ameliorate some of the cellular and phenotypic deficiencies. Since XPG is involved in multiple DNA repair systems, various types of DNA lesions will accumulate with age that differ in rates and outcome between distinct organs (27). We therefore primarily investigated the effect of vitamin E on the consequences of DNA damage in liver and nervous system.
The liver plays an important role in maintaining body homeostasis via the processing of nutrients, hormones, and metabolic waste products and by metabolic (in)activation of toxic compounds. It undergoes substantial changes in structure, composition and function in old age, thereby affecting systemic aging and disease predisposition (28, 29). One of the age-related changes in liver induced upon DNA damage, is the ability of hepatocytes to undergo polyploidization (4n, 8n, 16n and higher) (30). The level of polypoidization thereby shows a correlation with accumulation of DNA damage (14, 31). Vitamin E protected against polyploidization in a dose-dependent manner (Figure 3c and d). A logical explanation would be that increasing the amount of vitamin E via dietary changes would, via its anti-oxidant properties, prevent ROS from accumulating and damaging the DNA. Since direct measurements of oxidative DNA lesions are very challenging and often variable we investigated some intracellular intermediates in between lipids and DNA. Although protein carbonylation, a footprint of ROS damage associated with many neurodegenerative diseases and progeria (32), was not reduced by the slight physiological increase in the high vitamin E diet, likely due to the relatively high levels of other vitamins and anti-oxidants as feed preservatives, overall redox levels, reflecting a more general antioxidant change, showed a shift towards a more reduced state upon increasing the amount of vitamin E (Figure 3a and b). This however does not rule out alternative protective functions of vitamin E (18).
In brain, p53 is a ubiquitously expressed transcription factor activated by multiple types of DNA damage and therefore often used as biomarker for neurodegeneration (11). P53 analysis showed that consumption of feed with high vitamin E content was sufficient to reduce the number of p53 positive cells throughout the brain suggesting a neuroprotective role for vitamin E (Figure 5). Of note, no difference was observed when a diet deprived in vitamin E was compared with the diet containing medium vitamin E levels. This is in agreement with the vitamin E measurements between liver and brain of mice on the low vitamin E diet. It also substantiates the importance of adequate coverage of the organism with vitamin E and that, under increased oxidative pressure, a high vitamin E diet is needed to reduce the degenerating changes. Without further optimization, the current dosage of vitamin E supplementation did not yield a subsequent beneficial effect in astrogliosis in the brain or cell death in the retina (Figure 4). However, the reduced liver polyploidization and lower activation of the DNA damage response marker p53, indicates that higher levels of vitamin E might prevent or reduce age-dependent DNA damage accumulation.
Vitamin E is a lipid-soluble compound whose anti-oxidant properties have been investigated during aging progression or in pathologies typical of old age (17, 20, 21). The difficulty in performing precise and uniform human studies resides in several important variables such as: age and health status of the individuals included in the study, nutritional status and diet, form of vitamin E administered, but also quantity and time of supplementation. Other important parameters to be considered are also the genetic variations that can alter the vitamin E metabolism in the tissues as it was demonstrated by Goncalves and collaborators (33). Therefore, to investigate whether the observed and above mentioned preventive effects were indeed associated with alterations in vitamin E, we examined whether the orally supplemented α-Tocopherol was concentrated in the targeted organs studied. We measured vitamin E concentration in plasma, liver and brain of all experimental animals and confirmed that the diet enriched in vitamin E increased the concentration of this micronutrient in the analyzed tissue, while the diet with low vitamin E corresponded to vitamin E deficiencies (Table 1). Of note, brain showed the smallest reduction in vitamin E in animals subjected to the low vitamin E diet. These data confirm that the brain expresses high levels of the alpha-tocopherol transfer protein resulting in an enrichment and consequentially higher levels of vitamin E in the brain [34]. Moreover, our data corroborate what previously reported by Ulatowski and collaborators (35), who reported that vitamin E plays an important role in protecting the central nervous system, and specifically the cerebellar cortex by oxidative status with consequential deficits in motor coordination and cognitive functions following vitamin E deficiencies. Additionally, our findings demonstrate that a relatively brief period of vitamin E shortage leads to almost complete depletion of vitamin E from the tissue of these animals, emphasizing the importance for an elevated vitamin E supplementation under at-risk conditions. Other studies show that, in normal rats under vitamin E-deficient diet, a longer delay was observed for vitamin E concentrations to change in tissues studied (36, 37). The reason for this apparent discrepancy may be that in those studies lower levels of vitamin E were used and that we are dealing with a mouse model of premature aging in which the phenotype is fast developing. This may lead to an increased rate of metabolic processes and differences in vitamin E accumulation. Moreover, and in agreement to our results also in those studies the vitamin E supplementation led to marked elevation of its concentrations in different organs [36]. In addition, we measured the concentration of other micronutrients whose metabolism could have been affected by vitamin E content. Vitamin A, vitamin B1, B2 and B6 did not change in plasma, liver and brain as a function of the administered diet (Table 2) and therefore confirmed that vitamin E was the only variable explaining the observed parameters.
In conclusion, the data suggest that a diet rich in vitamin E delays the onset of specific aging features in the Xpg-/- premature aging model. In particular, we showed that a slight physiological increase in vitamin E reduces the number of damaged cells in liver and brain possibly in part due to an overall reduction of the oxidative stress status in these tissues. Our findings are of great potential for the implementation of focused therapeutic protocols in patients and imply the importance of improved nutrition to sustain health.


Acknowledgments: We thank Yvette van Loon, Lucien Rooth and the animal caretakers for general assistance with mouse experiments. Pier Giorgio Mastroberardino and Chiara Milanese are acknowledged for their support with the oxidative stress measurements. We acknowledge financial support of the National Institute of Health (NIH)/National Institute of Ageing (NIA) (PO1 AG017242), European Research Council Advanced Grant DamAge and Proof of Concept Grant Dementia to JHJH, the European commission ITN Address (GA-316390) and ITN Marriage (GA-316964), the KWO Dutch Cancer Society (5030), the Dutch CAA Foundation and the Royal Academy of Arts and Sciences of the Netherlands (academia professorship to JHJH). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Author contributions: G.L.F., J.H.J.H., W.P.V., and M.H.M. designed the research and wrote the manuscript. N.V., S.B., R.M.C.B., S.E., E.C., C.G., N.S. and P.W. contributed to editing the manuscript. S.B., R.M.C.B., and W.P.V. performed and analyzed the mouse cohorts and phenotypical scoring. S.E., E.C., and C.G., performed the quantitative vitamin measurements. N.V., S.B., and W.P.V. quantified oxidative stress parameters. N.V. performed the analysis of liver polyploidization and fat accumulation. N.V. and R.M.C.B. assessed the degree of retina degeneration. G.L.F. and W.P.V. characterized neuropathological changes.

Conflict of interests: G.L.F, S.E., E.C., C.G., N.S., P.W., and M.H.M. are employees of DSM Nutritional Products. All other authors declare no conflict of interests.

Statement of ethical approval: This study was performed in strict accordance with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and was approved by the Dutch Ethical Committee (permit # 139-12-13), in full accordance with European legislation.

Supplementary Figure

Supplementary Figure


1.     Lopez-Otin, C., et al., The hallmarks of aging. Cell, 2013. 153(6): p. 1194-217.
2.    Vermeij, W.P., J.H. Hoeijmakers, and J. Pothof, Genome Integrity in Aging: Human Syndromes, Mouse Models, and Therapeutic Options. Annu Rev Pharmacol Toxicol, 2016. 56: p. 427-45.
3.    von Zglinicki, T., A. Burkle, and T.B. Kirkwood, Stress, DNA damage and ageing — an integrative approach. Exp Gerontol, 2001. 36(7): p. 1049-62.
4.    Hoeijmakers, J.H., DNA damage, aging, and cancer. N Engl J Med, 2009. 361(15): p. 1475-85.
5.    Dan Dunn, J., et al., Reactive oxygen species and mitochondria: A nexus of cellular homeostasis. Redox Biol, 2015. 6: p. 472-85.
6.    Marteijn, J.A., et al., Understanding nucleotide excision repair and its roles in cancer and ageing. Nat Rev Mol Cell Biol, 2014. 15(7): p. 465-81.
7.    Mu, D., D.S. Hsu, and A. Sancar, Reaction mechanism of human DNA repair excision nuclease. J Biol Chem, 1996. 271(14): p. 8285-94.
8.    Klungland, A., et al., Base excision repair of oxidative DNA damage activated by XPG protein. Mol Cell, 1999. 3(1): p. 33-42.
9.    Trego, K.S., et al., Non-catalytic Roles for XPG with BRCA1 and BRCA2 in Homologous Recombination and Genome Stability. Mol Cell, 2016. 61(4): p. 535-46.
10.    Zafeiriou, D.I., et al., Xeroderma pigmentosum group G with severe neurological involvement and features of Cockayne syndrome in infancy. Pediatr Res, 2001. 49(3): p. 407-12.
11.    Barnhoorn, S., et al., Cell-autonomous progeroid changes in conditional mouse models for repair endonuclease XPG deficiency. PLoS Genet, 2014. 10(10): p. e1004686.
12.    Harkema, L., S.A. Youssef, and A. de Bruin, Pathology of Mouse Models of Accelerated Aging. Vet Pathol, 2016. 53(2): p. 366-89.
13.    Solon-Biet, S.M., et al., The ratio of macronutrients, not caloric intake, dictates cardiometabolic health, aging, and longevity in ad libitum-fed mice. Cell Metab, 2014. 19(3): p. 418-30.
14.    Vermeij, W.P., et al., Restricted diet delays accelerated ageing and genomic stress in DNA-repair-deficient mice. Nature, 2016. 537(7620): p. 427-431.
15.    Kamal-Eldin, A. and L.A. Appelqvist, The chemistry and antioxidant properties of tocopherols and tocotrienols. Lipids, 1996. 31(7): p. 671-701.
16.    Rigotti, A., Absorption, transport, and tissue delivery of vitamin E. Mol Aspects Med, 2007. 28(5-6): p. 423-36.
17.    Brigelius-Flohe, R. and M.G. Traber, Vitamin E: function and metabolism. FASEB J, 1999. 13(10): p. 1145-55.
18.    Mocchegiani, E., et al., Vitamin E-gene interactions in aging and inflammatory age-related diseases: implications for treatment. A systematic review. Ageing Res Rev, 2014. 14: p. 81-101.
19.    La Fata, G., P. Weber, and M.H. Mohajeri, Effects of vitamin E on cognitive performance during ageing and in Alzheimer’s disease. Nutrients, 2014. 6(12): p. 5453-72.
20.    Mohajeri, M.H. and G. Leuba, Prevention of age-associated dementia. Brain Res Bull, 2009. 80(4-5): p. 315-25.
21.    Mohajeri, M.H., B. Troesch, and P. Weber, Inadequate supply of vitamins and DHA in the elderly: implications for brain aging and Alzheimer-type dementia. Nutrition, 2015. 31(2): p. 261-75.
22.    Holler, U., et al., Microwave-assisted rapid determination of vitamins a and e in beverages. J Agric Food Chem, 2003. 51(6): p. 1539-42.
23.    Vermeij, W.P., A. Alia, and C. Backendorf, ROS quenching potential of the epidermal cornified cell envelope. J Invest Dermatol, 2011. 131(7): p. 1435-41.
24.    Horowitz, M.P., et al., Single-cell redox imaging demonstrates a distinctive response of dopaminergic neurons to oxidative insults. Antioxid Redox Signal, 2011. 15(4): p. 855-71.
25.    Vermeij, W.P. and C. Backendorf, Reactive oxygen species (ROS) protection via cysteine oxidation in the epidermal cornified cell envelope. Methods Mol Biol, 2014. 1195: p. 157-69.
26.    Li, S., et al., The Role of Oxidative Stress and Antioxidants in Liver Diseases. Int J Mol Sci, 2015. 16(11): p. 26087-124.
27.    Vermeij, W.P., J.H. Hoeijmakers, and J. Pothof, Aging: not all DNA damage is equal. Curr Opin Genet Dev, 2014. 26: p. 124-30.
28.    Le Couteur, D.G., et al., Old age and the hepatic sinusoid. Anat Rec (Hoboken), 2008. 291(6): p. 672-83.
29.    Hilmer, S.N., et al., Age-related changes in the hepatic sinusoidal endothelium impede lipoprotein transfer in the rat. Hepatology, 2005. 42(6): p. 1349-54.
30.    Duncan, A.W., et al., The ploidy conveyor of mature hepatocytes as a source of genetic variation. Nature, 2010. 467(7316): p. 707-10.
31.    White, R.R., et al., Controlled induction of DNA double-strand breaks in the mouse liver induces features of tissue ageing. Nat Commun, 2015. 6: p. 6790.
32.    Stadtman, E.R. and B.S. Berlett, Reactive oxygen-mediated protein oxidation in aging and disease. Drug Metab Rev, 1998. 30(2): p. 225-43.
33.    Goncalves, A., et al., Cluster-determinant 36 (CD36) impacts on vitamin E postprandial response. Mol Nutr Food Res, 2014. 58(12): p. 2297-306.
34.    Gohil, K., et al., Mice lacking alpha-tocopherol transfer protein gene have severe alpha-tocopherol deficiency in multiple regions of the central nervous system. Brain Res, 2008. 1201: p. 167-76.
35.    Ulatowski, L., et al., Vitamin E is essential for Purkinje neuron integrity. Neuroscience, 2014. 260: p. 120-9.
36.    Clement, M. and J.M. Bourre, Graded dietary levels of RRR-gamma-tocopherol induce a marked increase in the concentrations of alpha- and gamma-tocopherol in nervous tissues, heart, liver and muscle of vitamin-E-deficient rats. Biochim Biophys Acta, 1997. 1334(2-3): p. 173-81.
37.    Gaedicke, S., et al., Dietary vitamin E, brain redox status and expression of Alzheimer’s disease-relevant genes in rats. Br J Nutr, 2009. 102(3): p. 398-406.



S.-P. Chan1,2,3,4, P.Z.Yong5, Y. Sun5, R. Mahendran5,6,7, J.C.M. Wong5,6, C. Qiu8, T.-P. Ng5, E.-H. Kua5,6, L. Feng5


1. Department of Medicine, Yong Loo Lin School of Medicine, National University of Singapore; 2. Department of Decision Sciences, NUS School of Business, National University of Singapore; 3. Cardiovascular Research Institute, National University Heart Centre Singapore; 4. School of Engineering & Mathematical Sciences, La Trobe University, Australia; 5. Department of Psychological Medicine, Yong Loo Lin School of Medicine, National University of Singapore; 6. Department of Psychological Medicine, National University Hospital; 7. Duke-NUS Medical School, Singapore; 8. Aging Research Center, Department of Neurobiology, Care Sciences and Society, Karolinska Institutet-Stockholm University, Stockholm, Sweden

Corresponding Author: Dr Lei Feng, Department of Psychological Medicine, National University Health System Tower Block Level 9, 1E Kent Ridge Road, Singapore 119228
Republic of Singapore, Email:

J Prev Alz Dis 2018;5(1):21-25
Published online June 13, 2017,



Objective: To examine the association between long-term tea consumption and depressive and anxiety symptoms in community-living elderly.
Design: Community based cross-sectional study.
Setting: The Diet and Healthy Aging Study (DaHA), a prospective cohort study in Singapore.
Participants: 614 elderly aged 60 years and above, who were free of dementia and cognitive impairment.
Measurements:  Information on tea consumption was obtained through interviewer-administered questionnaire.   Long-term tea drinking was defined as regular consumption for at least 15 years. Depressive and anxiety symptoms were measured using the 15-item Geriatric Depression Scale (GDS-15) and the 20-item Geriatric Anxiety Inventory (GAI), respectively. A generalized structural equation model (gSEM) was applied to ascertain the association between long-term tea consumption and depressive and anxiety symptoms.
Results: About 59% of the subjects had consumed tea for over 15 years. Long term tea consumption was significantly associated with a reduced odds of having depressive and anxiety symptoms, after adjusting for demographics (i.e., age, gender, education and ethnicity), comorbid conditions (i.e., heart disease, diabetes, stroke, hypertension and hyperlipidaemia) and long-term coffee consumption.
Conclusion: There was evidence suggesting that long-term tea consumption was associated with reduced depressive and anxiety symptoms among community-living elderly. This suggests that it is worthwhile to further investigate the role of tea’s bioactive compounds in promoting mental health in aging.

Key words: Tea, aging, depression, anxiety, generalized structural equation model.



The potential health benefits of tea consumption are well-documented (1-6). Recent studies have linked tea consumption with better mental health in aging (7-10). The neuroprotective effects of tea consumption could be attributed to its antioxidant and anti-inflammatory properties (4).  Tea’s antioxidant property is primarily contributed by tea catechins, theaflavins and thearubigins (11, 12). Tea leaves also contain L-theanine, a unique amino acid, which may promote human brain functions (13). While the short-term central nervous system stimulating effects of tea have been reported in literature (14), the association between long-term consumption of tea and elderly’s psychological health has not been established. From a life-course perspective, the duration of healthy dietary habits could provide a more consistent measure of health-related behaviour and accumulated exposure. Although the importance of anxiety in late-life mental health is gaining much attention in recent years (15), no observational studies examining the long-term anxiolytic effects of tea have been reported to date.
The aim of this study was to ascertain the association between long-term regular consumption of tea (>15 years) and the depressive and anxiety symptoms in community-living elderly in Singapore using baseline data from the Diet and Healthy Aging Study (DaHA) cohort.




The study subjects were community-living elderly from the ongoing Diet and Healthy Aging Study (DaHA) cohort.  Commenced in July 2011, DaHA aims to study the relationship between Asian diets and health among community-living elderly in Singapore. Singapore citizens and permanent residents aged 60 years and above and residing in Jurong, an urban community in the western region of the city-state, were recruited by trained research staff through door-to-door visits. All interviews and assessments were conducted at the Training and Research Academy at Jurong Point (TaRA@JP), which is located in the targeted community. The protocol of DaHA was approved by the National University of Singapore Institutional Review Board, and written informed consent was obtained from all participants.

Data Collection

Detailed information on tea consumption was collated through an interviewer-administered questionnaire. The questions were designed according to the habitual consumption patterns among the elderly in Singapore. The common tea types were green tea, Chinese oolong tea, and English black tea. The key question on tea consumption was “How often do you consume each of the following food or drink?” at enrolment and at the age of 45, with six options for frequency, namely i. never or rarely, ii. more than once a month but less than once a week, iii. one to three times a week, iv. four to six times a week, v. one to two times a day, and vi. three times or more a day. With this one could determine whether the subjects had consumed tea for 15 years or longer (0: No, 1: Yes) given that they were at least 60 years old when enrolled. A subject was considered a long-term drinker (i.e., >15 years) if s/he had reported consuming tea or coffee regularly, at the age of 45 and at enrolment. Note that both tea and coffee are common beverages among the Singapore population.
The participants’ level of depression was measured by the 15-item Geriatric Depression Scale (GDS-15) (16, 17). The severity of anxiety was ascertained using the 20-item Geriatric Anxiety Inventory (GAI) (18). For both scales, a higher score is reflective of a greater severity of the symptoms. Other relevant information such as socio-demographics (e.g., age, gender, ethnicity, marital status, educational attainment and housing type) and health/medical conditions (e.g., heart disease, stroke, diabetes, hypertension, and hyperlipidaemia) were self-reported through the interviewer-administered questionnaire. A subject was considered to have suffered from heart disease if s/he had reported heart failure, heart attack or irregular heart-beat (atrial fibrillation).

Statistical Analysis

The sample characteristics were presented in median/range for quantitative variables, and in frequency/percentage for qualitative variables. Exploratory analyses were carried out with Wilcoxon-Mann-Whitney test and chi-square test, depending on the nature of data. In view of the above-mentioned study aim, the multivariate generalized structure equation model (gSEM) (19) was applied for confirmatory analyses. With the admissible values of GDS and GAI bounded between 0-15 and 0-20 respectively, a continuous and bounded Beta distribution was applied as the underlying distributions for the two scores. The flexible Beta distribution could handle data that are symmetrical or skewed (20). Both the GDS and GAI were standardized based on the standard version of Beta distribution and according to the conventional practice without loss of information (21). Long-term tea consumption (>15 years vs. ≤ 15 years or non-drinker) and the presence of heart disease, stroke, diabetes, hypertension and hyperlipidaemia followed the Binomial distribution in view of their binary nature (0: Without, 1: With). In both instances, the link functions applied were logit, thus enabling the estimated coefficients be interpreted as adjusted odds ratios (AOR). A backward elimination with removal probability of 0.05 was executed to identify the final model, but the non-significant predictors could be retained if they were relevant for answering the key questions of the study. The final chosen model is depicted in Figure 1 in the form of a path diagram, which indicates all hypothesized directionality of the analysis. For example, Tea→GDS indicates that an analysis would be performed for ascertaining long-term tea consumption’s effect on GDS-15 scores. A robust procedure was also implemented for correcting the standard errors in anticipation of outliers. The association between long-term coffee consumption and depression and anxiety was analyzed using the same approach for comparison and adjustment. The analysis was performed using Stata MP version 14 (Stata Corporation, Texas, USA), and all statistical tests were carried out at 5% level of significance.



Of the total of 642 elderly subjects enrolled from 2011 to 2015, 614 (95.6%) were free from dementia and cognitive impairment (MMSE ≥ 24). Of these, 97.2% were Chinese, 69.7% were women, 70.2% were married and 67.5% had primary or no formal education. While relatively few reported with heart disease (9.6%) and stroke (3.6%), a large number of subjects had suffered from hypertension (48.2%) and hyperlipidaemia (50.8%).  About 15% reported to have diabetes. The sample characteristics are depicted in Table 1.

Table 1. Characteristics of the study sample

Table 1. Characteristics of the study sample

Of the 614 subjects aged between 60 and 93, 362 (59.0%) reported to have consumed tea for more than 15 years.  While there was no significant age, ethnicity and housing type differences detected, more male subjects were long-term tea consumers when compared with their female counterparts (65.0% vs. 56.1%; p: 0.04).  There were also more elderly subjects with at least secondary school qualifications who consumed tea on a long-term basis, when compared with those who had lower educational attainment (70.1% vs. 53.5%; p<0.01).

Table 2. Final confirmatory analyses of GDS and GAI with gSEM

Table 2. Final confirmatory analyses of GDS and GAI with gSEM

*Statistically significant at 5%.

The GDS-15 (median: 1; range: 0-11) and GAI (median: 0; range: 0-17) scores were positively skewed, as a total of 248 subjects (40.4%) did not present any symptoms of depression while 433 (70.5%) were free from anxiety. Confirmatory analysis with gSEM showed that subjects who had consumed tea for over 15 years were more likely to have lower scores for GDS-15 (AOR: 0.82, p=0.01) and GAI (AOR: 0.84, p<0.01), while adjusting for their demographics, comorbid conditions and long-term coffee consumption. However, long-term coffee consumption was not significantly associated with depressive symptoms and anxiety symptoms.
A further analysis with gSEM showed that age, gender and ethnicity were not significantly associated with the prevalence of diabetes, stroke, hypertension and hyperlipidaemia, but heart disease was more prevalent in Malay and Indian male subjects in their advanced years.  Long-term tea and coffee consumption were also not significantly associated with the prevalence of heart disease, diabetes, stroke, hypertension and hyperlipidaemia.

Figure 1. Path diagram of the final gSEM

Figure 1. Path diagram of the final gSEM



The study found that long-term tea consumption was associated with a lower level of depression and anxiety among community living elderly in Singapore.  The findings provide fresh evidence on the potential role of tea and its bioactive compounds in promoting mental health in the aging population.
Tea’s antioxidant property is contributed by catechin, theaflavins and thearubigins. As a nonselective A1 and A2α adenosine receptor antagonist, caffeine could stimulate cholinergic neurons (22, 23) and help to prevent amyloid-β induced cognitive deficits (24).  Tea leaves also contain L-theanine, a unique amino acid, which may promote human brain functions and has been found to have significant impact on mental state (13).
The findings on depression replicated our previous studies using data from the Singapore Longitudinal Aging study (10) and the Confucius Hometowns Aging Project (9).  The inverse association between tea consumption and depressive symptom suggests that tea drinking could be a potential preventive measure for promoting mental health at the population level. It reinforces the evidential support concerning tea’s impact on preventing depression, as reported in literature (25, 26). But our findings concerning the association between long-term tea consumption and prevention of anxiety is novel. As such, the potential anxiolytic impact of tea should be further studied in clinical trials with biological markers of tea intake.
We used gSEM in statistical analysis because it is ideal for dealing with complex data structures. As the latest member of the SEM family of statistical techniques, gSEM is most suitable for dealing with and testing for complex inter-relationships among variables of mixed-types (qualitative and quantitative). A loss of precision (i.e., inflated standard errors) is expected should the data be analysed with the conventional models which could only deal with one outcome at a time, thus failing to handle the entire data structure in one single analytical setting and providing the much-needed comprehensive picture of the data patterns.
Our study has a few limitations. First, the reported statistically-significant association among variables are not to be interpreted as cause and effect in view of the cross-sectional study design. Second, depression and anxiety were measured using self-reported questionnaires but not clinical assessment and consensus diagnosis based on established criteria such as the DSM and ICD.  Last but not least, the study was not designed to provide a compelling answer to tea’s anti-pathogenic effects on depression and anxiety.
As such, a longitudinal study design is recommended for future studies. In fact, we are currently conducting a 5-year follow up of the DaHA cohort in Singapore, in order to facilitate the testing of dose-response relationship and to ascertain the effects of different types of tea (i.e., green tea, oolong tea and black tea).


Funding: Dr Lei Feng has received grant supports from the Virtual Institute for the Study of Aging, National University of Singapore (VG-8), the Alice Lim Memorial Fund, Singapore (Alice Lim Award 2010), the National Medical Research Council of Singapore (NMRC/TA/0053/2016), and an international grant from the Swedish Research Council for Health, Working Life and Welfare (Project No.: 2015-01352).

Acknowledgements: The research team would like to express its sincere gratitude to Lee Kim Tah Holdings Ltd., Singapore, Kwan Im Thong Hood Cho Temple, Singapore and the Presbyterian Community Services, Singapore, for their support.

Conflict of interest:  None of the authors reported potential conflict of interest.

Ethical standards: The study was approved by the National University of Singapore Institutional Review Board. All participants gave informed consent before taking part.



1.     Song J, Xu H, Liu F, Feng L. Tea and Cognitive Health in Late Life: Current Evidence and Future Directions. J Nutr Health Ageing 2012;16:31-34.
2.    Peters U, Poole C, Arab L. Does Tea Affect Cardiovascular Disease? A Meta-Analysis. Am J Epidemiol 2001;154:495-503.
3.    Liu G, Mi XN, Zheng XX, Xu YL, Lu J, Huang XH. Effects of tea intake on blood pressure: a meta-analysis of randomised controlled trials. Brit J Nutr 2014;112:1043-1054.
4.    Arab L, Liu W, Elashoff D. Green and Black Tea Consumption and Risk of Stroke: A Meta-Analysis. Stroke 2009;40:1786-1792.
5.    Ruan R, Feng L, Li J, Ng TP, Zeng Y. Tea consumption and mortality in the oldest-old Chinese. J Am Geriatr Soc 2013;61:1937-1942.
6.    Ng TP, Aung KC, Feng L, Feng L, Nyunt MS, Yap KB. Tea consumption and physical function in older adults: a cross-sectional study. J Nutr Health Ageing 2014;18:161-166.
7.    Feng L, Gwee X, Kua EH, Ng TP. Cognitive function and tea consumption in community dwelling older Chinese in Singapore. J Nutr Health Ageing 2010;14:433-438.
8.    Feng L, Chong MS, Lim WS, Gao Q, Nyunt MS, Lee TS, Collinson SL, Tsoi T, Kua EH, Ng TP. Tea Consumption Reduces the Incidence of Neurocognitive Disorders: Findings from the Singapore Longitudinal Aging Study. J Nutr Health Ageing 2016;20:1002-1009.
9.    Feng L, Yan Z, Sun B, Cai C, Jiang H, Kua EH, Ng TP, Qiu C. Tea consumption and depressive symptoms in older people in rural China. J Am Geriatr Soc 2013;61:1943-1947.
10.    Feng L, Li J, Kua EH, Lee TS, Yap KB, John Rush A, Ng TP. Association between tea consumption and depressive symptoms in older Chinese adults. J Am Geriatr Soc 2012;60:2358-2360.
11.    Inoue M, Tajima K, Hirose K, Hamajima N, Takezaki T, Kuroishi T, Tominaga S. Tea and coffee consumption and the risk of digestive tract cancers: data from a comparative case-referent study in Japan. Cancer Cause Control 1998;9:209-216.
12.    Leung LK, Su Y, Chen R, Zhang Z, Huang Y, Chen ZY. Theaflavins in black tea and catechins in green tea are equally effective antioxidants. J Nutr 2001;131:2248-2251.
13.    Nobre AC, Rao A, Owen GN. L-theanine, a natural constituent in tea, and its effect on mental state. Asian Pac J Clin Nutr 2008;17 Suppl 1:167-168.
14.    Smith A. Effects of caffeine on human behavior. Food Chem Toxicol 2002;40:1243-1255.
15.    Wolitzky-Taylor KB, Castriotta N, Lenze EJ, Stanley MA, Craske MG. Anxiety disorders in older adults: a comprehensive review. Depress Anxiety 2010;27:190-211.
16.    Sheikh JI, Yesavage JA. Geriatric Depression Scale (GDS): recent evidence and development of a shorter version. In clinical gerontology: a guide to assessment and intervention, 1986. The Haworth Press, New York, pp.165-173
17.    Broekman BF, Nyunt SZ, Niti M, Jin AZ, Ko SM, Kumar R, Fones CS, Ng TP. Differential item functioning of the Geriatric Depression Scale in an Asian population. J Affect Disorders 2008;108:285-290.
18.    Pachana NA, Byrne GJ, Siddle H, Koloski N, Harley E, Arnold E. Development and validation of the Geriatric Anxiety Inventory. Int Psychogeriatr 2007;19:103-111.
19.    Skrondal A, Rabe-Hesketh S. Generalized latent variable modeling: multilevel, longitudinal, and structural equation models. 2001. Chapman & Hall/CRC, Boca Raton
20.    Gupta AK, Nadarajah S. Handbook of beta distribution and its applications. 2004. Marcel Dekker, New York
21.    Smithson M, Verkuilen J. A better lemon squeezer? Maximum-likelihood regression with beta-distributed dependent variables. Psychol Methods 2006;11:54-71.
22.     Fredholm BB, Bättig K, Holmén J, Nehlg A, Zvartau EE. Actions of caffeine in the brain with special reference to factors that contribute to its widespread use. Pharmacol Rev 1999;51:83-133.
23.     Eskelinen MH, Ngandu T, Tuomilehto J, Soininen H, Kivipelto M. Midlife coffee and tea drinking and the risk of late-life dementia: a population-based CAIDE study.  J Alzheimers Dis 2009;16:85-91.
24.     Dall’Igna OP, Fett P, Gomes MW, Souza DO, Cunha RA, Lara DR. Caffeine and Adenosine A2β receptor antagonists prevent beta-amyloid (25-35)-induced cognitive deficits in mice. Exp Neurol 2007;203:241-245.
25.    Niu K, Hozawa A, Kuriyama S, Ebihara S, Guo H, Nakaya N, Ohmori-Matsuda K, Takahashi H, Masamune Y, Asada M, Sasaki S, Arai H, Awata S, Nagatomi R, Tsuji I. Green tea consumption is associated with depressive symptoms in the elderly. Am J Clin Nutr 2009;90:1615–1622.
26.    Zeng Y, Chan HS, Ni T, Ruan RP, Feng L, Nie C, Cheng LG, Li Y, Tao W, Gu J, Land KC, Yashin A, Tan QH, Yang Z, Bolund L, Yang HM, Hauser E, Willcox CD, Willcox BJ, Tian XL, Vaupel JW. GxE interactions between FOXO genotypes and tea drinking significantly affect cognitive disability at advanced ages in China. J Gerontol A Biol Sci Med Sci 2015;70(4):426-33.



R.F. Buckley1,2,3,4, K.P. Sparks1,5, K.V. Papp1,2,5, M. Dekhtyar1,5, C. Martin6, S. Burnham7, R.A. Sperling1,2,5, D.M. Rentz1,2,5


1. Massachusetts General Hospital, Boston, Massachusetts, USA; 2. Harvard Medical School, Boston, Massachusetts, USA; 3. Florey Institutes of Neuroscience and Mental Health, Melbourne, Australia; 4. Melbourne School of Psychological Sciences, University of Melbourne, Australia; 5. Brigham and Women’s Hospital, Boston, Massachusetts, USA; 6. Northeastern University, Boston, Massachusetts, USA; 7. Commonwealth Scientific and Industrial Research Organization, Perth, Australia

Corresponding Author: Dorene M. Rentz,  Harvard Medical School,  Department of Neurology,  Brigham and Women’s Hospital,  60 Fenwood Road,  Boston, MA 02115,  Phone 617-732-8235,  Email:

J Prev Alz Dis 2017;4(1):3-11
Published online January 24, 2017,



Background: As prevention trials for Alzheimer’s disease move into asymptomatic populations, identifying older individuals who manifest the earliest cognitive signs of Alzheimer’s disease is critical. Computerized cognitive testing has the potential to replace current gold standard paper and pencil measures and may be a more efficient means of assessing cognition. However, more empirical evidence about the comparability of novel computerized batteries to paper and pencil measures is required.
Objectives: To determine whether two computerized IPad batteries, the NIH Toolbox Cognition Battery and Cogstate-C3, similarly predict subtle cognitive impairment identified using the Preclinical Alzheimer Cognitive Composite (PACC).
Design, Setting, Participants: A pilot sample of 50 clinically normal older adults (Mage=68.5 years±7.6, 45% non-Caucasian) completed the PACC assessment, and the NIH Toolbox and Cogstate-C3 at research centers of Massachusetts General and Brigham and Women’s Hospitals. Participants made 3-4 in-clinic visits, receiving the PACC first, then the NIH Toolbox, and finally the Cogstate-C3.
Measurements: Performance on the PACC was dichotomized by typical performance (>= 0.5SD), versus subtle cognitive impairment (<0.5SD). Composites for each computerized battery were created using principle components analysis, and compared with the PACC using non-parametric Spearman correlations. Logistic regression analyses were used to determine which composite was best able to classify subtle cognitive impairment from typical performance.
Results: The NIH Toolbox formed one composite and exhibited the strongest within-battery alignment, while the Cogstate-C3 formed two distinct composites (Learning-Memory and Processing Speed-Attention). The NIH Toolbox and C3 Learning-Memory composites exhibited positive correlations with the PACC (ρ=0.49, p<0.001; ρ=0.58, p<0.001, respectively), but not the C3 Processing Speed-Attention composite, ρ=-0.18, p=0.22. The C3 Learning-Memory was the only composite that classified subtle cognitive impairment, and demonstrated the greatest sensitivity (62%) and specificity (81%) for that subtle cognitive impairment.
Conclusions: Preliminary findings suggest that the NIH Toolbox has the advantage of showing the strongest overall clustering and alignment with standardized paper-and-pencil tasks. By contrast, Learning-Memory tasks within the Cogstate-C3 battery have the greatest potential to identify cross-sectional, subtle cognitive impairment as defined by the PACC

Key words: Cognition, Neuropsychology, Aging, Computerized Testing.



Interest in using computerized cognitive testing as a potential outcome measure in clinical trials has steadily increased. Computerized testing has been proposed as a feasible and reliable way of testing older participants (1-4). Studies examining the validity of computerized cognitive composites in relation to performance on conventional neuropsychological instruments are accruing (5-8), and furthermore, computerized testing has already become a secondary outcome in a major clinical trial (9). Until recently, however, clinical trials have relied upon conventional paper and pencil neuropsychological tests, as they represent a gold-standard in clinical testing and diagnostic decision-making (for a discussion, see: 10). As technology advances, clinical trials are increasingly moving towards validated computerized testing for sensitively capturing cognitive performance in large-scale secondary prevention cohorts. Comparing computerized batteries against current measures used in large-scale clinical trials is critical as the field moves towards these large-scale, population-based cognitive screening and assessments (11, 12). The Alzheimer’s Disease Cooperative Study Preclinical Alzheimer Cognitive Composite (PACC) (9, 13) is a composite of standard paper and pencil tests that are currently being used in a large-scale prevention trial (9). The PACC was originally designed as a multi-domain but memory-predominant cognitive composite that exhibited sensitivity to biomarker risk of AD in clinically-normal older adults (13).
It is unclear how computerized batteries perform in relation to one another against conventional paper-and-pencil composites, such as the PACC. Secondly, it is not clearly understood how these batteries may compare in their ability to classify subtle cognitive impairment as defined by poor performance on paper-and-pencil composites. Two computerized batteries that are of particular relevance to these questions are the Cogstate Computerized Cognitive Composite (C3) battery (1), which is currently being used in the Anti-Amyloid Treatment in Asymptomatic Alzheimer’s Disease (A4) secondary prevention trial (9), and the newly developed, non-proprietary iPad version of the National Institutes of Health (NIH) Toolbox Cognition Battery (NIHTB-CB) (for reference to the general computerized battery: 14, 15). The Cogstate C3 departs from the original Cogstate Brief Battery (7) as it includes the Face-Name Associative Memory Exam (FNAME), a challenging associative memory task found to be sensitive to neocortical amyloid burden in older adults (16), and the Behavioral Pattern Separation Task-Object (BPXT) (17), a pattern-separation memory task sensitive to treatment change in an MCI trial (17). The Cogstate Brief Battery is well-validated, and has been shown to capture AD-related cognitive changes in older adults (18), and those with MCI and AD (19). The desktop version of the NIHTB-CB has been validated against standard neuropsychological measures, and in a large and demographically diverse population ranging in age from 3 to 85 years (6, 14). The NIHTB-CB is intended to serve as a ‘common currency’ among longitudinal and epidemiological studies, however, it is yet to be tested in clinical trials or longitudinal observational studies of aging and dementia. Neither of these batteries are a direct replication or ‘digitization’ of conventional paper-and-pencil tests, but represent a novel approach to cognitive testing that that can be optimally translated to computerized technologies. As an example, Cogstate utilizes playing cards as a non-verbal assessment of working memory and processing speed that has wide cross-cultural applicability (18).
A critical component of early detection in preclinical Alzheimer’s disease is the ability of neuropsychological tests to identify evidence of subtle cognitive impairment (20). Defined as Stage 3, after abnormal levels of both amyloidosis and neurodegeneration are apparent, the appearance of subtle changes in cognitive performance heralds the final phases of preclinical AD prior to a diagnosis of MCI. Targeting clinically-normal older adults at risk of AD-related cognitive decline over short term follow-up will require sophisticated cognitive batteries that are sensitive to subtle change, but will also need to meet the requirements of large-scale clinical trials in clinically-normal cohorts for being deployable across large populations. Before computerized batteries can be utilized in these environments, these batteries must demonstrate validity for identifying preclinical levels of subtle cognitive impairment (21, 22).
The aims of this pilot cross-sectional study were three-fold. First, we developed aggregate cognitive composites for both computerized batteries to measure overall cognitive performance in relation to the paper and pencil PACC. We also aimed to compare each of these computerized batteries against performance on the PACC in clinically normal older adults. Finally, using the PACC to define subtle cognitive impairment, our objective was to determine the ability of each of the computerized batteries to distinguish subtle cognitive impairment from typical cognitive performance. Evidence that these batteries similarly identify subtle cognitive impairment would support the validity of these instruments for large-scale screening and cognitive outcome protocols for clinical trials.


Materials and Methods


Fifty clinically normal, community-dwelling, older adults (age range: 54-97 years) were recruited from volunteers interested in research studies at the Center for Alzheimer Research and Treatment at Brigham and Women’s Hospital and at the Massachusetts Alzheimer Disease Research Center at Massachusetts General Hospital. All subjects underwent informed consent procedures approved by the Partners Human Research Committee, the Institutional Review Board for Brigham and Women’s Hospital and Massachusetts General Hospital. No prior computer or iPad knowledge was required. Subjects were excluded if they had a history of alcoholism, drug abuse, head trauma or current serious medical or psychiatric illnesses. All subjects met the age requirement (above 50 years old), and scored within age-specified norms on the Telephone Interview of Cognitive Status (TICS; 23). We set a minimum age of 50 years, as longitudinal research studies and clinical trials are beginning to include younger ages in their cohorts (i.e. the Australian Imaging Biomarker and Lifestyle (AIBL) study of ageing, the Harvard Aging Brain Study (HABS), and the ante-amyloid (A3) clinical trial (11)).


In order to mimic a typical clinical trial setting, subjects participated in three-four clinic visits within a six-month time-frame, where they completed the PACC, the NIHTB-CB, and the Cogstate iPad C3 battery at the first, second and third visit, respectively. Each visit was separated from the next by approximately one week. The rationale for multiple clinic visits was to reduce cognitive fatigue when completing each neuropsychological battery. Both computerized batteries were performed from beginning to end in one visit. Participants made a fourth visit as part of a larger study that will not be covered in the current study. Participants were not extensively trained to use the iPad prior to testing, as the tests were overseen by an examiner according to a standardized administration (CM, KPS, MD). Instructions were given if the participant was having trouble making selections (pressing too hard or too long).


The PACC includes Logical Memory–delayed recall (LM-DR), the Free and Cued Selective Reminding Test (FCSRT) total score, the Mini Mental State Exam (MMSE) total score, and Wechsler Adult Intelligence Scale-Revised Digit Symbol Coding Test (DSC) total score (13). This composite includes measures of general cognition (MMSE) and speeded executive function (DSC), but is 50% composed of episodic memory tests (13). All tests were z-transformed using the mean and standard deviation of performance by clinically normal older adults (n=256, age range: 61-90) years) participating in the Harvard Aging Brain Study (24, 25). This population served as an ideal normative sample by which to classify our current pilot sample as individuals were recruited from the same geographic area and recruited through the same centers. To form the PACC, all z-transformed variables were averaged together, with a higher score indicating better performance.
The NIHTB-CB included the Flanker Inhibitory Control and Attention Test (Flanker), the Picture Sequence Memory Test (PSMT), the Picture Vocabulary Test (PVT), the Pattern Comparison Processing Speed Test (PCPST) and the Dimensional Change Card Sort Test (DCCS) (14). Two other NIHTB-CB measures, the List Sorting Working Memory test and the Oral Reading Recognition test, were not included in the current study as they required the use of an additional keyboard. The Flanker is a measure of cognitive control, where the participant is asked to attend to a stimulus that is flanked by four identical stimuli that are either positioned congruently or incongruently to the target. The participant must select the direction in which the target stimulus is pointing. The PSMT is a measure of episodic memory in which participants are shown a series of images and asked to re-create the image order over two trials. The PVT is a measure of receptive vocabulary; participants are orally presented a word and are asked to select from one of four images that is closest to the meaning of the word. The outcome measure for PVT was age-scaled and standardized. The PCPST is a measure of processing speed, where participants are asked to match an object with response items by either color or shape. The DCCS is a measure of set shifting, where a participant matches a target visual stimulus to one or two choice stimuli according to shape or color (14). The PSMT, PCPST, DCCS and Flanker tasks were all computed scores provided by NIHTB-CB. These computed scores reflect a theta score, which reflects an individual’s overall ability or performance, similar to a z-score.
The Cogstate C3 includes the FNAME and the Behavioral Pattern Separation-Object Task (BPXT), as well as the Detection Task (DET), the Identification Task (IDN), the One Card Learning Task (OCL) and the One-Back Task (ONB). The FNAME is an associative memory test that requires participants to associate (FNMT), and subsequently recall (FNLT), and recognize (FSBT) faces with corresponding names. The FNAME task measures frequency of correct responses. The BPXT assesses working and recognition memory; participants are iteratively presented with a series of repeated, novel and distractor images and are asked to categorize each into Old, Similar, or New. The outcome measure is frequency of correct responses. Additional tasks use playing cards as stimuli. The DET is a measure of reaction time and processing speed, where participants are asked to respond when a stimulus card is turned face up. The IDN is an attention paradigm in which a card is presented and the respondent must choose whether the card is red or is not red (black). The outcome measures for these two tasks were speed (sec:ms). The OCL task is a non-verbal memory task, which assesses short term recall of a set of repeated playing cards. OCL was measured using accuracy. The ONB task is a measure of working memory, where respondents are asked to serially match each card to the previous trial, and was also measured according to speed of response (18). These scores were not transformed, however, they were converted to z-scores.

Creating computerized battery composites

Our initial aim was to create cognitive composites from the computerized batteries in order to align with the PACC. Previous studies have created cognitive composites from the Cogstate Brief battery in older adults who were clinically-normal and patients with MCI and AD, so we investigated whether the Cogstate C3 could create similar composites. NIHTB-CB Crystallized Cognition Composites and Fluid Cognition Composites have been proposed in a previous study, however, these were created from a sample of children and required two extra tests that we did not include in our study. Global composite measures were created for each of the NIHTB-CB and Cogstate C3 batteries using principal component analysis (PCA), and Bartlett factor scores were extracted. These composites were created consistent with previous reports using the Cogstate Brief Battery (19) and the NIHTB-CB (15). PCA was used to reduce the NIHTB-CB and C3 into global composite scores. Using scree plots and eigenvalue cut-offs, we determined that the NIHTB-CB could be reduced to one composite, while the C3 exhibited a better fit with two composites. The NIHTB-CB composite accounted for 47% of the variance explained in the model, while the two C3 composites accounted for a total of 61% of the variance (with the first factor accounting for 32% variance). The first C3 factor, ‘Learning-Memory’, included the BPXT, FNMT, FNLT, FSBT and OCL. The second factor, ‘Processing Speed-Attention’, included the ONB, IDN and DET. A clustering model using a two-dimensional PCA, which compared the similarities of the tasks in two-dimensional space was also used to explore how tasks clustered together. The results, displayed in Figure 1, suggested that IDN, DET and ONB created a distinct cluster, while the remainder of the NIHTB-CB and C3 tasks formed a second, tight cluster (see Figure 1). Using composite scores arising from this data reduction approach, allowed us to pursue our main hypotheses, i.e., using the composite scores to differentiate high and low performance in relation to the PACC.

Figure 1. Visualization of clusters using PCA of NIHTB-CB with C3 tasks. Arrows indicate the loading coefficients of each variable of interest

Figure 1. Visualization of clusters using PCA of NIHTB-CB with C3 tasks. Arrows indicate the loading coefficients of each variable of interest

Statistical methodology

Analyses were conducted using IBM SPSS version 22.0, and R (version 3.3.0). Due to a smaller sample size, a series of non-parametric Spearman correlations were conducted to ascertain relationships between computerized battery composites and PACC performance. Performance on the PACC was dichotomized into normal and subtle cognitive impairment according to a cut-off of 0.5SD below the normative group mean, which was derived according to the Harvard Aging Brain Study cohort (an entirely separate cohort from the one used in the current study). These participants were not considered to meet criteria for a diagnostic classification of mild cognitive impairment, but demonstrated a very subtle cognitive decrement in comparison with their peers. This classification was chosen to align with Stage 3 Preclinical AD criteria, which states: “Evidence of subtle cognitive decline, that does not meet criteria for MCI or dementia” (20). Choosing a 0.5SD cut-off allowed us to define subtly poorer performance, while maintaining samples with enough power for analytical purposes. This sits in contrast to diagnostic classifications for MCI which typically note performance below 1-1.5SD age-adjusted norms (26). Three logistic regression analyses were performed to determine how well the NIHTB-CB and C3 composites could detect the subtle cognitive impairment group. Although age and education levels were not found to be significantly different between typical PACC performers and those with subtle cognitive impairment, we ran analyses with these covariates included in order to portray our results within the context of age and education-adjustment. Receiver Operating Characteristic (ROC) curves were computed to determine sensitivity, specificity and area under the curve (AUC) parameters of each composite to classify normal and subtle cognitive impairment. Post-hoc analyses were run to ascertain which tests within the best performing composites were driving better classification outcomes.



Participant Characteristics

No demographic differences were found between those who were classified as subtle cognitive impairment or exhibiting typical performance according to the PACC (see Table 1). However, a marginally greater number of non-Caucasian individuals (n = 9) were found to be classified as subtly impaired on the PACC in comparison with Caucasians (n = 4), χ2(1) = 3.91, p = .10, however this difference was not statistically significant. The non-Caucasian group was not found to be significantly different from the Caucasian group on any demographics, although, there was a trend for lower education levels (ranging from 12-20 years), χ2(4) = 8.13, p = .09.

Table 1. Participant characteristics and cognitive performance

Table 1. Participant characteristics and cognitive performance

Note: Subtle cognitive impairment is PACC performance below 0.5SD *Cell sizes are too small to count

Associations between computerized batteries and PACC performance

The NIHTB-CB and C3 Learning-Memory were both associated with the PACC (ρNIHTB-CB(47) = 0.49, p < .001 and ρC3 Learning-Memory(47) = 0.58, p <.001). There was no significant relationship found between the PACC and C3 Processing Speed-Attention, ρ(47) = -0.18, p =.22.

Ability of computerized tasks to distinguish subtle cognitive impairment according to the PACC

Logistic regression analyses showed that the NIHTB-CB and Cogstate C3 Learning-Memory models were significantly able to distinguish subtle cognitive impairment from typical PACC performance, and explained 9% and 49% of variance in their respective models (χ2NIHTB-CB(42) = 48.22, p = .04 and χ2COGSTATE-C3(42) = 23.61, p < .001; see Table 2 for all three model fits and estimates). Greater NIHTB-CB performance related to better classification of those with subtle cognitive impairment, however, this finding did not survive multiple comparisons (B (SE) = 0.79 (0.4), p = .05). Better Learning-Memory performance significantly increased the chance of being classified with typical (i.e., better) PACC performance (B (SE) = 3.71 (1.2), p = .003). Our findings showed the same pattern of results with or without age and education included in the models (see Table 2).


Table 2. Regression and ROC analyses with each computerized composite to predict subtle cognitive impairment or typical performance on the PACC

Table 2. Regression and ROC analyses with each computerized composite to predict subtle cognitive impairment or typical performance on the PACC

Note: The large confidence intervals in this analysis are driven largely by the sample size, and so the OR should be interpreted with caution

ROC curves showed that performance on the C3 Learning-Memory composite accounted for the largest AUC (92%), and exhibited the greatest sensitivity (61%) and specificity (80%) indices for classifying subtle cognitive impairment (see Table 2 for all sensitivity and specificity parameters). Figure 3 depicts a scatterplot between the NIHTB-CB and Cogstate C3 (by averaging performance on both C3 composites) according to the PACC groups. Scores sitting in the top right-hand quadrant depict high performance on both computerized batteries; all but one of these scores included individuals with typical PACC performance, illustrating high specificity.
As the C3 Learning-Memory composite exhibited the highest odds ratio and ROC parameters, we ran a post-hoc logistic regression to determine which measures within the Learning-Memory composite (FNMT, FNLT, FSBT, BPXT and OCL) were driving these results. Better performance on the Face Name Letter Task, a measure of delayed free recall, was the only measure within the Learning-Memory composite found to significantly increase the likelihood of typical PACC performance (B (SE) = 5.6 (3.1), p = .05). As a comparison, we also conducted a post-hoc analysis with the NIHTB PSMT task, a free recall memory task, and found that better PSMT performance significantly predicted typical PACC performance, (OR = 3.3, p = .04, CI95%: 1.3-12.5). Neither the FNLT task nor the NIHTB PSMT task were better able to classify subtle cognitive impairment in comparison with the full composite measures, with AUC, sensitivity and specificity parameters comparable to their counterparts (see Table 2 and Figure 2).

Figure 2. ROCs for the NIHTB-CB and Cogstate C3 composites, and the C3 FNLT task alone to distinguish between high and low PACC performance (Blue = C3 Learning-Memory, Red = NIHTB-CB, Green = C3 Processing Speed-Attention, Grey = C3 FNLT, Black-dash = NIH PSMT)

Figure 2. ROCs for the NIHTB-CB and Cogstate C3 composites, and the C3 FNLT task alone to distinguish between high and low PACC performance (Blue = C3 Learning-Memory, Red = NIHTB-CB, Green = C3 Processing Speed-Attention, Grey = C3 FNLT, Black-dash = NIH PSMT)


Figure 3. Scatterplot of association between NIHTB-CB and Cogstate C3 battery, with slopes estimating group effect of high and low PACC performance

Figure 3. Scatterplot of association between NIHTB-CB and Cogstate C3 battery, with slopes estimating group effect of high and low PACC performance


This pilot study in normal older adults sought to directly compare performance on computerized batteries, the NIHTB-CB and Cogstate C3 batteries, to the PACC, a clinical trial outcome measure composed of conventional paper and pencil cognitive tasks. The Learning-Memory composite from the Cogstate C3 battery was able to distinguish between normal PACC performance and subtle cognitive impairment (see Fig 4 for a diagrammatic representation of findings). The composite also showed particularly high specificity and AUCs for correctly classifying normal individuals. These findings were found to be primarily driven by the delayed free recall index from the Face-Name task that was featured within the composite. By contrast, the NIHTB-CB yielded a moderate level of specificity, with a sensitivity at chance level, while the C3 Processing Speed-Attention composite was poor on both parameters. We did find, however, that the NIHTB-CB showed a comparable level of correlation with the PACC as was found with C3 Learning-Memory. By contrast, the C3 Processing Speed-Attention composite did not show any affinity with the PACC. This supports other findings suggesting that processing speed and attention domains are less sensitive to AD-related change very early in the trajectory (18), and perhaps are more sensitive to age-related etiologies (27). These results most likely reflect the nuanced differences in ‘intended purpose’ for the NIHTB-CB and Cogstate (C3) batteries. The NIHTB-CB has been proposed as a well-validated measure that can be utilized in a broad range of age-groups and education levels (14), while the Cogstate C3 is a battery primarily intended for clinical trials, and which has been shown to be sensitive to AD-related cognitive change (28).

Figure 4. Diagrammatic representation of each composite arising from the Cogstate C3 and NIHTH-CB computerized batteries, and their corresponding tests. Each composite is also attached to an odds ratio (OR) which represents the ability of each composite to distinguish between typical and subtly impaired PACC performance. The pink boxes denote the tasks that were most contributory to the variance explained in the logistic regression model

Figure 4. Diagrammatic representation of each composite arising from the Cogstate C3 and NIHTH-CB computerized batteries, and their corresponding tests. Each composite is also attached to an odds ratio (OR) which represents the ability of each composite to distinguish between typical and subtly impaired PACC performance. The pink boxes denote the tasks that were most contributory to the variance explained in the logistic regression model

One strength of the NIHTB-CB in this study was that it formed a clear singular composite, and displayed largely unified within-battery alignment as suggested by clustering methods. The NIHTB-CB has shown strong convergent validity with other standard neuropsychological paper-and-pencil tests along the broad developmental trajectory (14), and was originally designed to complement measures used in research studies of cognition or to serve as a brief adjunct measure in longitudinal and epidemiologic studies (14, 29). It was not, however, specifically developed as an early diagnostic tool for AD-related cognitive impairment or as a target for disease outcomes. The NIHTB composite was able to identify subtle cognitive impairment, particularly using the NIHTB memory task. This supports the notion that the NIHTB-CB is a suitable measure of cognitive performance in clinically-normal older adults. Sensitivity for classification of subtle cognitive impairment was not as high in comparison with the Cogstate C3 Learning-Memory composite. An additional advantage of the NIHTB-CB battery is that it includes a measure of IQ, which is not covered by the C3. As such, this battery has the unique potential to efficiently measure cognitive reserve outcomes, and may well have the ability to inform an individuals’ likely compensatory duration for increasing pathology over time. Our findings highlight the different possible utilities of these computerized batteries within the context of secondary prevention clinical trials. It is possible that the NIHTB-CB will be more sensitive to early longitudinal cognitive decline, however, the current pilot study is unable to investigate this question.
Within the Cogstate C3 battery, two distinct composites were extracted, similar to previous studies (18, 19, 30), supporting the notion that the Cogstate Battery was intended to measure distinct cognitive domains. The C3 Learning-Memory composite, however, showed an association with PACC performance, and an ability to classify subtle cognitive impairment. The Cogstate Brief battery has been shown to reliably highlight increasing magnitude of impairment in MCI and AD diagnostic groups, and that computerized performance tracks well with performance on conventional tests (18, 28). Our findings suggest that the FNAME component of the Cogstate C3 battery may be of particular interest for clinical trials of preclinical AD. Although evidence of subtle cognitive impairment was defined in our study, it is not solely an indication of stage 3 preclinical AD as we do not have indications of AD biomarker status. Furthermore, exhibiting subtle cognitive impairment does not by itself indicate progressive cognitive decline. As such, sensitivity to the classification of subtle cognitive impairment will need to be more fully determined by larger, longitudinal investigations. In addition, validation studies will be required in comparison populations of MCI and AD dementia. It may be that the ADAS-Cog and screening tools such as the MMSE are sufficient for clinical populations, but that more challenging neuropsychological tasks included in computerized batteries are more relevant for large-scale clinical trials of clinically-normal individuals. Our findings further suggest that not all C3 tasks have the ability to identify subtle cognitive decline, and as such, may not be necessary for inclusion in large-scale screening procedures for preclinical AD trials.
We found that the driving predictor of sensitivity to subtle cognitive impairment in the current study was the delayed free recall index from the C3 FNAME task. The Cogstate C3 departs from the Cogstate Brief Battery in that it includes the FNAME (1), which has been shown to be sensitive to amyloid-ß deposition (12). The addition of the FNAME measures in the C3 battery may have increased the ability of the C3 to capture variation in PACC performance, which is is the current standard for clinical trials (1). As the PACC is a composite that is more heavily weighted towards memory (by including two memory measures), and is honed to detect amyloid-related change (13), it is not surprising that memory components of the C3 battery are able to classify subtle impairment on the conventional composite. In the current study, delayed free recall from the FNLT was found to drive the group classification, which provides support for the recommendation that the FNAME be included in the Cogstate Brief Battery for longitudinal studies of memory in preclinical AD. Although it was a significant component of the composite to classify group performance, neither the C3 FNLT task nor the NIHTB PSMT task performed significantly better than their composite counterparts. While parsimonious neuropsychological batteries are advantageous, we currently recommend that full Cogstate Learning-Memory or NIHTB-CB batteries are performed.
The current study is a pilot study of clinically normal older adults, and as such we were limited to studying the classification of subtle cognitive impairment as defined by the PACC. Although, the sample size is small, the strength of this study is that it covers a broad range of older ages and maximizes the racial diversity of subjects. As no major demographic differences were present in typical and subtle cognitive impairment PACC performers, we did not covary for race in our analyses, although we acknowledge that more sophisticated examinations of diversity-related cognitive profiles should be conducted in larger samples (31). In addition, we did not acquire AD biomarkers, and cannot conclude on the extent to which these tests measure biological markers of interest. In the future, we plan to include the NIHTB-CB and C3 in a larger cohort of clinically normal older adults who have undergone AD biomarkers and intend to follow the performance of these individuals over time. In addition, it will be important to counterbalance for battery administration, and assess in-home compared to in-clinic testing performance. The trend is moving towards large-scale online cognitive testing, as evidenced by registries that include online testing such as the Brain Health Registry (32) and the UKBioBank (33). Determining test-retest reliability between at-home and in-clinic testing will be vital. Large secondary prevention trials that require access to trial-ready cohorts who are identified based on cognitive performance are needed. Computerized on-line testing, that is well validated, will make this feasible. We believe that both iPad batteries presented in this study, show promise as valid cognitive assessments in the clinical trial setting. However, more work will be needed before they can be effectively utilized as on-line cognitive tests for large-scale prevention trials.


Funding: Neurotrack Technologies funded this study. Rachel F. Buckley is funded by the NHMRC/ARC Dementia Research Fellowship (APP1105576). Reisa A. Sperling has served as a paid consultant for Abbvie, Biogen, Bracket, Genentech, Lundbeck, Merck, Otsuka, Roche, and Sanofi. She has served as a co-investigator for Avid, Eli Lilly, and Janssen Alzheimer Immunotherapy clinical trials. She has spoken at symposia sponsored by Eli Lilly, Biogen, and Janssen Alzheimer Immunotherapy. Dorene M. Rentz has served as a paid consultant for Eli Lilly, Lundbeck Pharmaceuticals and Biogen Idec. She also serves on the Scientific Advisory Board for Neurotrack. Kathryn V. Papp has served as a paid consultant for Biogen Idec. These relationships are not related to the content in the manuscript.

Acknowledgements: We would like to thank Drs. Sandy Weintraub, Jerry Slotkin, and Paul Maruff for their invaluable comments and input to the development of this manuscript.

Ethical standards: The Partners Human Research Committee approved this study. All subjects underwent informed consent.



1.    Rentz DM, Dekhtyar M, Sherman J, Burnham S, Blacker D, Aghjayan SL, Papp KV, Amariglio RE, Schembri A, Chenhall T. The Feasibility of At-Home iPad Cognitive Testing For Use in Clinical Trials. JPAD 2016;3, 8-12.
2.    Wild K, Howieson D, Webbe F, Seelye A, Kaye J. Status of computerized cognitive testing in aging: a systematic review. Alzheimer’s & Dementia 2008;4, 428-437.
3.    Sano M, Egelko S, Ferris S, Kaye J, Hayes TL, Mundt JC, Donohue M, Walter S, Sun S, Sauceda-Cerda L. Pilot Study to Show the Feasibility of a Multicenter Trial of Home-based Assessment of People Over 75 Years Old. Alzheimer disease and associated disorders 2010;24, 256-263.
4.    Fredrickson J, Maruff P, Woodward M, Moore L, Fredrickson A, Sach J, Darby D. Evaluation of the usability of a brief computerized cognitive screening test in older people for epidemiological studies. Neuroepidemiology 2009;34, 65-75.
5.    Steinberg SI, Negash S, Sammel MD, Bogner H, Harel BT, Livney MG, McCoubrey H, Wolk DA, Kling MA, Arnold SE. Subjective Memory Complaints, Cognitive Performance, and Psychological Factors in Healthy Older Adults. American Journal of Alzheimer’s Disease & Other Dementias 2013;28, 776-783.
6.    Weintraub S, Dikmen SS, Heaton RK, Tulsky DS, Zelazo PD, Slotkin J, Carlozzi NE, Bauer PJ, Wallner-Allen K, Fox N. The cognition battery of the NIH toolbox for assessment of neurological and behavioral function: Validation in an adult sample. JINS 2014;20, 567-578.
7.    Maruff P, Thomas E, Cysique L, Brew B, Collie A, Snyder P, Pietrzak RH. Validity of the CogState brief battery: relationship to standardized tests and sensitivity to cognitive impairment in mild traumatic brain injury, schizophrenia, and AIDS dementia complex. Archives of Clinical Neuropsychology 2009;24, 165-178.
8.    Hammers D, Spurgeon E, Ryan K, Persad C, Barbas N, Heidebrink J, Darby D, Giordani B. Validity of a brief computerized cognitive screening test in dementia. Journal of geriatric psychiatry and neurology 2012;25, 89-99.
9.    Sperling RA, Rentz DM, Johnson KA, Karlawish J, Donohue M, Salmon DP, Aisen P. The A4 Study: Stopping AD Before Symptoms Begin? Science Translational Medicine 2014;6, 228fs213.
10.    Bauer RM, Iverson GL, Cernich AN, Binder LM, Ruff RM, Naugle RI. Computerized Neuropsychological Assessment Devices: Joint Position Paper of the American Academy of Clinical Neuropsychology and the National Academy of Neuropsychology(). Archives of Clinical Neuropsychology 2012;27, 362-373.
11.    Sperling RA, Mormino E, Johnson K. The Evolution of Preclinical Alzheimer’s Disease: Implications for Prevention Trials. Neuron 2014;84, 608-622.
12.    Rentz DM, Parra Rodriguez MA, Amariglio R, Stern Y, Sperling RA, Ferris S.  Promising developments in neuropsychological approaches for the detection of preclinical Alzheimer’s disease: a selective review. Alzheimer’s Research & Therapy 2013;5, 58.
13.    Donohue MC, Sperling RA, Salmon DP, et al. The preclinical alzheimer cognitive composite: Measuring amyloid-related decline. JAMA Neurology 2014;71, 961-970.
14.    Weintraub S, Dikmen SS, Heaton RK, Tulsky DS, Zelazo PD, Bauer PJ, Carlozzi NE, Slotkin J, Blitz D, Wallner-Allen K. Cognition assessment using the NIH Toolbox. Neurology 2013;80, S54-S64.
15.    Akshoomoff N, Beaumont JL, Bauer PJ, Dikmen SS, Gershon RC, Mungas D, Slotkin J, Tulsky D, Weintraub S, Zelazo PD, Heaton RK. NIH Toolbox cognition battery (CB): Composite scores of crystallized, fluid, and overall cognition. Monographs of the Society for Research in Child Development 2013;78, 119-132.
16.    Rentz DM, Amariglio RE, Becker JA, Frey M, Olson LE, Frishe K, Carmasin J, Maye JE, Johnson KA, Sperling RA. Face-name associative memory performance is related to amyloid burden in normal elderly. Neuropsychologia 2011;49, 2776-2783.
17.    Stark SM, Yassa MA, Lacy JW, Stark CE. A task to assess behavioral pattern separation (BPS) in humans: Data from healthy aging and mild cognitive impairment. Neuropsychologia 2013;51, 2442-2449.
18.    Lim YY, Ellis KA, Harrington K, Ames D, Martins RN, Masters CL, Rowe C, Savage G, Szoeke C, Darby D. Use of the CogState Brief Battery in the assessment of Alzheimer’s disease related cognitive impairment in the Australian Imaging, Biomarkers and Lifestyle (AIBL) study. Journal of Clinical and Experimental Neuropsychology 2012;34, 345-358.
19.    Maruff P, Lim YY, Darby D, Ellis KA, Pietrzak RH, Snyder PJ, Bush AI, Szoeke C, Schembri A, Ames D. Clinical utility of the cogstate brief battery in identifying cognitive impairment in mild cognitive impairment and Alzheimer’s disease. BMC Psychology 2013;1, 30.
20.    Sperling RA, Beckett L, Bennett D, Craft S, Fagan A, Kaye J, Montine T, Park D, Reiman E, Siemers E, Stern Y, Yaffe K. Criteria for preclinical Alzheimer’s disease. Alzheimer’s Association, 2010.
21.    Carrillo MC, Vellas B. New and different approaches needed for the design and execution of Alzheimer’s clinical trials. Alzheimer’s & dementia: the journal of the Alzheimer’s Association 2013;9, 436-437.
22.    Reiman EM, Langbaum J, Fleisher AS, Caselli RJ, Chen K, Ayutyanont N, Quiroz YT, Kosik KS, Lopera F, Tariot PN. Alzheimer’s Prevention Initiative: a plan to accelerate the evaluation of presymptomatic treatments. Journal of Alzheimer’s Disease 2011;26, 321-329.
23.    Knopman DS, Roberts RO, Geda YE, Pankratz VS, Christianson TJ, Petersen RC, Rocca WA. Validation of the telephone interview for cognitive status-modified in subjects with normal cognition, mild cognitive impairment, or dementia. Neuroepidemiology 2010;34, 34-42.
24.    Dagley A, LaPoint M, Huijbers W, Hedden T, McLaren DG, Chatwal JP, Papp KV, Amariglio RE, Blacker D, Rentz DM. Harvard aging brain study: Dataset and accessibility. In press, 2015;S1053-8119.
25.    Hedden T, Mormino EC, Amariglio RE, Younger AP, Schultz AP, Becker JA, Buckner RL, Johnson KA, Sperling RA, Rentz DM. Cognitive profile of amyloid burden and white matter hyperintensities in cognitively normal older adults. The Journal of Neuroscience 2012;32, 16233-16242.
26.    Albert MS, DeKosky ST, Dickson D, Dubois B, Feldman HH, Fox NC, Gamst A, Holtzman DM, Jagust WJ, Petersen RC. The diagnosis of mild cognitive impairment due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimer’s & Dementia 2011;7, 270-279.
27.    Hedden T, Gabrieli JDE. Insights into the ageing mind: a view from cognitive neuroscience. Nat Rev Neurosci 2004;5, 87-96.
28.    Lim YY, Ellis KA, Harrington K, Pietrzak RH, Gale J, Ames D, Bush AI, Darby D, Martins RN, Masters CL, Rowe CC, Savage G, Szoeke C, Villemagne VL, Maruff P.  Cognitive Decline in Adults with Amnestic Mild Cognitive Impairment and High Amyloid-β: Prodromal Alzheimer’s Disease? J Alz Dis 2013;33, 1167-1176.
29.    Gershon RC, Wagster MV, Hendrie HC, Fox NA, Cook KF, Nowinski CJ. NIH toolbox for assessment of neurological and behavioral function. Neurol 2013;80, S2-S6.
30.    Lim YY, Ellis KA, Ames D, Darby D, Harrington K, Martins RN, Masters CL, Rowe CC, Savage G, Szoeke C, Villemagne VL, Maruff P. Aβ amyloid, cognition, and APOE genotype in healthy older adults. Alzheimer Dem 2013;9, 538-545.
31.    Manly JJ, Schupf N, Tang M-X, Stern Y.  Cognitive decline and literacy among ethnically diverse elders. Journal of Geriatric Psychiatry and Neurology 2005;18, 213-217.
32.    Nosheny RL, Flennkiken D, Insel PS, Finley S, Mackin S, Camacho M, Truran-Sacrey D, Maruff P, Weiner MW (2015) Internet-based recruitment of subjects for prodromal and secondary prevention Alzheimer’s disease trials using the brain health registry. Alzheimer’s & Dementia: The Journal of the Alzheimer’s Association 2015;11, P156.
33.    Matthews PM, Sudlow C. The UK Biobank. Brain 2015;138, 3463-3465.


C. Vassalle1, L. Sabatino2, A. Pingitore2, K. Chatzianagnostou1, F. Mastorci2, R. Ceravolo3

1. Fondazione CNR-Regione Toscana G Monasterio, Pisa, Italy; 2. Institute of Clinical Physiology-CNR, Pisa, Italy; 3. Neurology Institute-Department of Clinical and Experimental Medicine, University of Pisa, Pisa, Italy.

Corresponding Author: Dr. Cristina Vassalle, Fondazione G.Monasterio CNR-Regione Toscana, Pisa, Italy Phone: +39-050-3152199; Fax:+39-050-3153525. e-mail:

J Prev Alz Dis 2017;4(1):58-64
Published online August 16, 2016,


 This review aims to focus on main antioxidants- abundantly contained in the diet- as well as of the whole Mediterranean diet and life-style and their relationship with cognitive function, especially critical in two phases of life, in children until adolescence and oldness. The role of emerging biochemical and molecular biomarkers as opportunity to estimate more accurately nutritional assumption and requirement, in terms of cognitive preservation and disease risk, will be also discussed.  The cluster of factors within the Mediterranean pattern -which include not only nutritional, but also physical, social, and stimulating aspects- is still largely understudied as a whole, but it is proposed as attractive research area and tool for public health planning of prevention and intervention.

Key words: Nutrition, cognitive decline, aging, Mediterranean diet, Mediterranean life-style, antioxidants, biomarkers.


The progressive aging of western population is closely related to increased costs imposed by social-sanitary needs in over 65 subjects. Cognitive decline consists in a deterioration of cognitive function, characterized by increasing difficulties with cognitive processing speed, process information, language, think abstractly and problem solving, including aspects related to memory, executive functioning and reasoning (1). This condition may progress to dementia when the symptoms are severe enough to impact significantly with daily life (2). In this context, cognitive decline often related to progressive loss of personal independence and disability, will represent one of the main problem to face in the future society. Conversely, the knowledge of underlying mechanisms of brain morphology and function will constitute a critical step in developing preventive and therapeutic strategies that meet the demands of an aging society.
Normally, cognitive abilities increase up until adolescence or early adulthood and then they progressively decline. However, the age at which the decline begins and the speed of ability diminishes are highly variable among subjects (Figure 1) (3). Life-style habit, including nutrition and diet, can contribute to individual differences, and early life-phase living conditions may affect late-life health outcomes.

Figure 1. Functional capacity trend along life span and interval of intervention to maintain the best possible levels of brain functionality and prevent cognitive impairment and disability

Therefore, this review will focus on available evidences of single antioxidants or Mediterranean diet (MeD) and their effects on cognitive function in two critical phases of life: 1) from childhood to adolescence and 2) oldness. Moreover, since a low endogenous antioxidant status and inflammation may represent key factors for cognitive decline, the possible use of biomarkers related to these processes as tool to assess the relationship between diet and cognitive decline will be also evaluated. Finally, the cluster of factors within the Mediterranean style-life, still largely understudied as a whole including not only dietary, but also physical, social, and stimulating  factors, is proposed as attractive research area and tool for public health planning of prevention and intervention.

Nutrition in early life and later cognitive function

The effects of nutrition on cognitive function is crucial since pre-birth life and factors like the diet,  intake of folates or maternal vitamin B12 status during pregnancy or lactating have significant impact on successive measures of language, memory, and perception. However, there is a lack of pregnancy and birth cohorts to study ageing from the life-course perspective and to monitor how subjects age according to different biological as well as lifestyle variables (Table 1). In fact, different endpoints have been evaluated mostly in childhood, due to the  many difficulties arising to plan prospective studies with 10 years or more of follow-up (4-7). Moreover, often studies do not include subjects with frankly reduced nutrient status, and in most cases, the retrospective nature of data collection for food insufficiency in childhood may raise concern of recall bias.  In any case, some interesting observations may be extrapolated by available findings, suggesting that malnutrition during the first years of life drives risk for significant functional morbidity in adulthood. In fact, poor living conditions in early life may cause higher risk for chronic conditions such as depression, hypertension, Type 2 diabetes (T2D), and obesity, which in turn may drive to worse neuronal outcomes later during the adulthood. In this context, recent findings from a 40-year longitudinal study suggest that moderate to  severe malnutrition during infancy is associated with impaired IQ and academic skills in adulthood (8). Accordingly, very recent data revealed that food insufficiency in childhood would independently increase the risk of developing dementia in old age by 81%, after adjusting for sociodemographic factors (9). Conversely, results from the  1932 Scottish Mental Survey evidenced that lower B12 at age 79 is associated with cognitive decline between age 11 and 79, while serum folates at age 79 correlates with those at age 11 (10). On 2435 participants in the community-based Coronary Artery Risk Development in Young Adults (CARDIA) study of black and white men and women (18-30 years at the time of enrollment, 1985-86), the diet score was associated with cognitive function of the following 5 years and even 25 years, evidencing the importance of early adoption and maintenance of quality diet to preserve intellectual capacity along all life span (11).

Table 1. Nutrition and age-related cognitive impairment: actual limits and needs

Whether the results of these trials are not simple to perform and interpret in terms of underlying molecular mechanisms, some interesting insights may derive by experimental studies. Interestingly, very recent data suggest that improved cognition in adult rats -subjected to low calorie diet feeding during youth- may result from the increase of brain-derived neurotrophic factor (BDNF) involved in hippocampal neurogenesis (12). In particular, neuronal nuclear antigen-neuron marker expressing cells, which are involved in memory and are located in hippocampus dentate gyrus, resulted increased (12). Moreover, hippocampus and prefrontal cortex BDNF levels were increased, while serum glucose concentration and values of malondialdehyde (marker of lipid peroxidation) appear reduced in serum and hippocampus (12). Thus, BDNF could represent a critical underlying factor in the relationship between nutrition and cognition.

Antioxidants, diet and cognitive decline in elderly

Data from WHO predicted that the occurrence of cognitive impairment and dementia will interest 29 million people worldwide in 2020 (13). This alarming information is anyway alleviated by the prediction that even a small delay of the onset of Alzheimer disease (AD) may markedly reduce the disease prevalence and, consequently, even modest interventions postponing disease onset could translate in a major public health impact (14). In light of this consideration, many studies on the effects of single key nutrients -folate, vitamin B12, and vitamin E- in the elderly have been recently reviewed (15). Authors concluded that supplementation may protect against cognitive decline but only in elderly subjects with low status of these vitamins (for folate is <12 nmol/L or vitamin E intake <6.1 mg/day) (15). No clear definitive data emerge for vitamin B12 (15). Some other studies revealed an interaction between plasma concentrations of folate and vitamin B12 in relation to cognitive performance. In a large population of 2203 Norwegian elderly aged 72-74 years, where cognitive performance was assessed by six cognitive tests, vitamin B12 in the lowest quartile (< 274 pmol/L), combined with plasma folate in the highest quartile (>18.5 nmol/L) were associated with a reduced risk of cognitive impairment (16). Other results from the Framinghan cohort suggested that low vitamin B12 levels (<258 pmol/L) may predict cognitive decline, being a higher cognitive decline rate observed in subjects with low vitamin B12 and high plasma folate (>21.75 nmol/L) or supplemental folates (17).  In other studies, high folate or folic acid supplements resulted detrimental to cognition in older people with low vitamin B12 levels (18). Conversely, a recent study on the effects of 2-year folic acid and vitamin B12 supplementation on cognitive performance in elderly people with elevated homocysteine levels (2,919 elderly participants, ≥65 years) showed no change in cognitive performance (19). Nonetheless, in the Chicago Health and Aging Project (516 participants) higher levels of vitamin B12 were associated with slower rates of cognitive decline, although homocysteine concentration had no relationship to cognitive decline (20).
Recently, there has been increasing interest in the potential of flavanols, contained in fruit and vegetables, to improve cognitive functions in elderly. A high-flavanol intake was found to enhance dentate gyrus activity (hippocampal region considered involved in age-related memory decline), as measured by Functional Magnetic Resonance Imaging and cognitive tests (21). Moreover, these molecules are able to improve regional cerebral perfusion in elderly, which can be one of the possible acute mechanism by which flavanols exert their benefits on cognitive performance (22). Moreover, other data suggested that also the habitual consumption of green tea may be effective in reduce the risk of cognitive decline in elderly population (23, 24). Interestingly, thiamine (or vitamin B1 from cereals, meat, vegetables) deficiency is common in frail elderly (especially hospitalized and institutionalized subjects) and has been related to AD and other neurodegenerative conditions. (25-28). However, the role of thiamine on cognitive function in elderly subjects was recently reviewed, and authors concluded that at now there are no definitive data to clearly recognize the role of thiamine in neurological impairment and disease (28).
Although even the results on a single nutrient may be complex to understand, because often contradictory or associated to variables effect according to nutrient concentration, as in the case of vitamins and homocysteine, studies on the effects of a single food may be further limited because do not focus on the interactive effects with other components in a whole diet. In fact, intake is clearly based on a complex interaction of both macro- (proteins, fats, carbohydrates) and micro-nutrients (vitamins, minerals), and evaluation of the effects of a single nutrient or food could be not satisfactory enough. Actually, an interesting approach to examine the link between nutrition and cognitive function is the evaluation of whole dietary patterns, thus considering potential synergies among nutrients, as found in a balanced diet. In this context, both the Dietary Approaches to Stop Hypertension (DASH) and MeD dietary patterns were associated with consistently higher levels of cognitive function in elderly men and women over an 11-year period in a prospective design (29). These results were confirmed in the 826 Memory and Aging Project (826 participants, aged 81.5 ± 7.1 years) where both the DASH and MeD patterns resulted associated with slower rates of cognitive decline (30). The molecular mechanisms by which DASH and MeD may be related to beneficial effects on cognitive performance have not yet been fully cleared. However, the commonly diffused use of whole grains, nuts and legumes may be responsible for the similar protective effects of both DASH and MeD dietary patterns. This observation may be an important starting point for public-health nutrition recommendations worldwide, according to the wide diffusion of these food types in the diet of majority world populations.

MeD and cognitive decline in elderly

MeD is the food pattern typical of population living in Italy and other Mediterranean countries, characterized by a high fish, vegetable and  fruit consumption,  use of extravirgin  olive oil, low intake of dairy derivatives, sugar infrequent consume, moderate red wine use, red meat consumption (only 1-2 times/week). Noteworthy, a high adherence to MeD is associated with longevity and provides significant protection against morbidity and mortality related to chronic diseases including cancer, metabolic syndrome, depression, cardiovascular and neurodegenerative diseases (31).
There are also findings that a better adherence to MeD may reduce the risk of cognitive decline and some forms of dementia (32). In particular, a meta-analysis assessed the association between the MeD and Mild Cognitive Impairment (MCI) or AD from five prospective cohort studies with at least one year of follow up. Higher MeD adherence confers a reduced risk (33% lower) of developing both MCI and AD, and a reduced risk of progressing from MCI to AD (33). Another meta-analysis evaluated the association between MeD and a number of brain-related conditions, stroke, depression, and cognitive impairment (8 studies covered cognitive impairment), suggesting that high MeD adherence was strongly associated with reduced risk for cognitive impairment (MCI, dementia and AD) (34). These data were confirmed by other results from recent systematic revision and meta-analysis, evidencing that MeD adherence reduced the risk of developing MCI and AD, and the progression from MCI to AD (35-37). The interaction mechanism between MeD adherence and cognition could be almost in part linked to the oleocanthal, an extra-virgin olive-oil bioactive component, recently supposed to be involved in the modulation of tau protein, one of the main causes of Alzheimer neurodegeneration (38). Accordingly, greater adherence to MeD was associated with better scores in several cognitive function tests in elderly subjects (>60 years) living in a Polish rural community (39), thus suggesting some benefits even in non-Mediterranean populations when the main MeD principles were adopted. Conversely, adverse effects of ‘Western’ dietary patterns against the consumption of high vegetable and plant-based diet was evidenced in elderly (>60 years) included in the Australian Diabetes, Obesity and Lifestyle Study (40).

Oxidative stress and inflammatory biochemical markers in the relationship between MeD and cognitive decline

Biochemical markers represent indices of a biological state or condition, measurable in biological samples, especially into the blood. As antioxidants are major determinants of protective MeD effects, the measurements of inflammatory and oxidative stress biomarkers may contribute to fill the gap between nutrition and MeD on one side and age-related cognitive impairment and AD on the other. In particular, MeD adoption increases levels of carotenoids, vitamin A and vitamin E, and reduced oxidative stress and inflammation biomarker levels (e.g. uric acid, SH groups, SOD and GPx activities, FRAP and TRAP, TNF-α, and IL-10 cytokines, and malondialdehyde in the erythrocytes as marker of lipid peroxidation) (41). Recent data also suggested that MeD adherence favorably modifies levels of oxidative stress biomarkers (CoQ and β-carotene, isoprostanes and oxidized low-density lipoproteins) in elderly subjects (42). Some of these oxidative stress and inflammatory biomarkers have been proposed to evidence early risk AD profiles, even in the pre-symptomatic stage, and as additive tools for AD diagnosis, and prognosis (43-46). However, at now, there is scarcity of data on the possibility to modulate levels of inflammatory and oxidative stress biomarkers by nutrients and dietary patterns in patients with MCI or AD. In particular, there are not significant evidences for a role of C reactive protein, a common index of inflammation, in the association between MeD and lower risk of AD (118 incident AD cases during a 4-year follow-up in 1219 non-demented subjects aged  over 65 yrs) (47).
Beneficial effects of MeD have supposed to be related to prevention of shortening of telomeres, nucleoprotein structures that protect the ends of chromosomes, whose integrity is closely related to antioxidant availability (48-50). In this context, results from the PREDIMED-NAVARRA (PREvención con DIeta MEDiterránea-NAVARRA) study evidenced that diet significantly modulates telomere length, as an inverse relationship was observed between obesity parameters (body weight, body mass index, waist circumference and waist to height ratio) and telomere length (48). Telomeres may also retain potential value as risk biomarkers for MCI and AD, as shorter telomere length has been associated with several age-related chronic degenerative diseases (51, 52). Nonetheless, to the best of our knowledge there are no studies which evaluate telomere modulation by MeD in patients with cognitive decline.
The “case” of resveratrol is interesting, because this polyphenol (contained in grapes, some nuts and dried fruits, and red wine) exerts beneficial effects in in vitro models of neudegenerative diseases, including AD, Parkinson and Huntington’s disease, epilepsy, amyotrophic lateral sclerosis, although results on animal and patients are still lacking (53). In particular, its neuroprotective actions appear almost in part related to the  activation of  the sirtuins’ family member SIRT1 (35, 54-56). Increased SIRT1 expression has been related to antioxidant upregulation, and downregulation of pro-apoptotic factors through the involvement of Forkhead box O transcription factors (Fox01-06) and oxidative stress reduction (35). Moreover, SIRT1 negatively regulates p66Shc, that increases intracellular ROS levels through an oxidoreductase activity, and NF-κB, whose DNA-binding capacity is decreased by deacetylation of its RelA/p subunit by SIRT1 (35). Thus, SIRT1, with these multiple effects, has been related to different critical pathways, regarding metabolic control, DNA repair, apoptosis, cell survival, development, inflammation, and mitochondrial function (35). Other components of the MeD can also induce SIRT1, such as the polyphenol quercetin contained in red wine and red onions (57), and polyphenols in extra virgin olive oil (58). Thus, all these nutrients and the MeD as a whole deserve to be evaluated in future studies as safe, no-pharmacological SIRT1 activating tools, able to elicit endogenous neural protection, and preventing MCI and AD.
Interestingly, the emerging interaction between genes and dietary intake may account for the complex interplay between individual shape and external environment, and supports the concept of variable individual response to nutrition and heterogeneity in cognitive ageing. Accordingly, the benefits on lipid profile of fish oil fatty acids eicosapentaenoic and docosahexaenoic acids consumption appeared dependent on Apolipoprotein E genotypes in the 2340 subjects enrolled in the Multi-Ethnic Study of Atherosclerosis (mean age 61±10 yrs) (59).  Moreover, cognitive decline in T2D appears associated with an altered metabolomic profile involving sphingolipids, bile acids, and uric acid metabolism (60). Very interesting experimental data showed how maternal folate depletion and high-fat feeding from weaning may affect DNA methylation and DNA repair in brain of adult mouse, causing a high sensibility to oxidative damage (61). Nutritional interventions in the adulthood could positively counteract epigenetic changes, including DNA methylation and microRNA, associated with ageing (62). However, recent data, obtained on 9-week-old C57BL/6J mice exposed to a high-fat diet for 15 weeks, showed the development of diet-induced obesity and insulin resistance with the onset of irreversible epigenetic modifications in the brain, which persisted also if normal metabolic homeostasis is restored (63). In this context, results from the NU-AGE multidisciplinary consortium of 30 partners from 17 European Union countries are expected. This study aims to evaluate whether a one-year Mediterranean whole diet can affect  physical and cognitive status in the elderly (65-79 years of age), utilizing biomarkers obtained by a series of analyses, including omics (transcriptomics, epigenetics, metabolomics and metagenomics) (64).

Mediterranean life-style and cognitive decline in elderly

The MeD was recognized by UNESCO in 2010 as a cultural heritage of Humanity (65). Although nutritional elements represent the core of the Mediterranean lifestyle, several other aspects must be considered, such as para-dietetic (gastronomy, preparation, setting) and meta-dietetic (cultural) features of MeD, which impact on health and well-being (66). In particular, the Mediterranean Diet Foundation has evidenced additive aspects other than food, such as conviviality, socialization, biodiversity and seasonality, and moderate physical activity as essential complement to the MeD dietary pattern (66). Specifically, frugality and moderation in food consumption characterized this life-style. There is biodiversity and use of seasonal, traditional food, short production and distribution line. In the traditional MeD life-style, family has lunch together at home, food was consumed slowly, with a glass of red wine, conviviality and conversation ensured, together with periods of calm and mid-day rest. Outdoor living included walking, gardening and raising their own vegetables. Moreover, they cook their own food for their own pleasure and others’ satisfaction. The conviviality aspect of eating is important, as it contributes to increase communication and socialization, whereas dedicate time to food preparation and intensify multisensory stimulation through tactile, gustative, and visual stimuli are also involved.  This cluster of factors related in the MeD life-style has not yet been evaluated in association to cognitive function and various diseases. However, it is known that these parameters are individually important to cognitive function and may interact synergistically. Socialization and physical activity resulted associated to a better cognitive performance (67).Very recent studies address mealtimes, seen as an opportunity for social interaction, which may improve health and behavior in elderly people (68-70). Food and physical activity are two parameters closely interrelated, which represent complementary aspects of the energy balance that have regulated brain development and function during human evolution. The combination of an healthy dietary approach together with appropriate physical activities during life-time can contribute to maintain or decrease cognitive impairment (71). Accordingly, dietary factors, such as a diet poor in saturated fat, and exercise have been evidenced as important determinants of cognitive performance in experimental and human studies, modulating neuronal and behavioral plasticity through the increase of BDNF levels (72, 73).


Although there are still main pitfalls (Table 1), available data suggested that amount, content of food, meal frequency and context might represent a non-pharmacological, low-cost, and low-tech interventional option, effective as environmental inducer of brain plasticity for the prevention and improvement of cognitive function across all life span.
The possible use of oxidative and inflammatory biomarkers to assess the relationship between diet and cognitive impairment and neurological disease is also promising, although a substantial amount of work remains in terms of replicating the few findings already available. The rapid development of metabolomics and other “omics” techniques may increase the chance of finding relevant parameters for a more reliable and accurate assessment of association of food intake with functional status and disease risk profile markers. In the future, the identification and use of nutritional biomarkers could allow to estimate more accurately nutritional assumption and requirement, taking into account bioavailability and individual differences, and opening innovative approaches for personalized nutritional plan of cognitive decline prevention.
More important, scientific and clinical medical advances had created a profound evolution in the concept of health, that according to WHO is defined as “a state of complete physical, mental and social well-being and not merely the absence of disease or infirmity”. In this context, the concept of synergy between factors within the Mediterranean pattern -which include not only nutritional, but also physical, social, and stimulating aspects- offers the opportunity to evaluate the impact of a comprehensive lifestyle cluster. Such holistic approach may help to develop more efficacious interventional strategies  promoting wellbeing across the life span and preventing disabilities and cognitive impairment  beyond the biological, psychological and social barriers that aging implies.

Conflict of interest: None



1.    Hardman RJ, Kennedy G, Macpherson H, Scholey AB, Pipingas A. A randomized controlled trial investigating the effects of Mediterranean diet and aerobic exercise on cognition in cognitively healthy older people living independently within aged care facilities: the Lifestyle Intervention in Independent Living Aged Care (LIILAC) study protocol (ACTRN12614001133628). Nutr J 2015;14:53.
2.    Vega JN, Newhouse PA. Mild cognitive impairment: diagnosis, longitudinal course, and emerging treatments. Curr Psychiatry Rep. 2014;16:490.
3.    Hartshorne JK, Germine LT. When does cognitive functioning peak? The asynchronous rise and fall of different cognitive abilities across the life span. Psychol Sci 2015;26:433-43.
4.    Burkhalter TM, Hillman CH. A narrative review of physical activity, nutrition, and obesity to cognition and scholastic performance across the human lifespan. Adv Nutr 2011;2:201-6S.
5.    Bhate V, Deshpande S, Bhat D, Joshi N, Ladkat R, Watve S, et al. Vitamin B12 status of pregnant Indian women and cognitive function in their 9-year-old children. Food Nutr Bull 2008;29:249-54.
6.    Bonilla C, Lawlor DA, Taylor AE, Gunnell DJ, Ben-Shlomo Y, Ness AR, et al. Vitamin B-12 status during pregnancy and child’s IQ at age 8: a Mendelian randomization study in the Avon longitudinal study of parents and children. PLoS One 2012;7:e51084.
7.    Villamor E, Rifas-Shiman SL, Gillman MW, Oken E. Maternal intake of methyl-donor nutrients and child cognition at 3 years of age. Paediatr Perinat Epidemiol 2012;26:328-35.
8.    Waber DP, Bryce CP, Girard JM, Zichlin M, Fitzmaurice GM, Galler JR. Impaired  IQ and academic skills in adults who experienced moderate to severe infantile malnutrition: a 40-year study. Nutr Neurosci. 2014;17:58-64.
9.    Momtaz YA, Haron SA, Hamid TA, Ibrahim R, Masud J. Does food insufficiency in  childhood contribute to dementia in later life? Clin Interv Aging 2014;10:49-53.
10.    Starr JM, Pattie A, Whiteman MC, Deary IJ, Whalley LJ. Vitamin B-12, serum folate, and cognitive change between 11 and 79 years. J Neurol Neurosurg Psychiatry 2005;76:291-2
11.    Zhu N, Jacobs DR, Meyer KA, He K, Launer L, Reis JP, et al. Cognitive function in a middle aged cohort is related to  higher quality dietary pattern 5 and 25 years earlier: the CARDIA study. J Nutr Health Aging 2015;19:33-38.
12.    Kaptan Z, Akgün-Dar K, Kapucu A, Dedeakayoğulları H, Batu Ş, Üzüm G. Long term consequences on spatial learning-memory of low-calorie diet during adolescence in female rats; hippocampal and prefrontal cortex BDNF level, expression of NeuN and cell proliferation in dentate gyrus. Brain Res 2015 Jun 10. (doi: 10.1016/j.brainres.2015.05.041. PubMed PMID: 26072462)
13.    Haan MN, Wallace R. Can dementia be prevented? Brain aging in a population-based context. Annu Rev Public Health 2004;25:1–24.
14.    Brookmeyer R, Gray S, Kawas C. Projections of Alzheimer’s disease in the United States and the public health impact of delaying disease onset. Am J Public Health 1998;88:1337-42.
15.    Barnes JL, Tian M, Edens NK, Morris MC. Consideration of nutrient levels in studies of cognitive decline. Nutr Rev 2014;72:707-19.
16.    Doets EL, Ueland PM, Tell GS, Vollset SE, Nygård OK, Van’t Veer P, et al. Interactions between plasma concentrations of folate and markers of vitamin B(12) status with cognitive performance in elderly people not exposed to folic acid fortification: the Hordaland Health Study. Br J Nutr 2014;111:1085-95.
17.    Morris MS, Selhub J, Jacques PF. Vitamin B-12 and folate status in relation to decline in scores on the mini-mental state examination in the framingham heart study. J Am Geriatr Soc 2012;60:1457-64.
18.    Moore EM, Ames D, Mander AG, Carne RP, Brodaty H, Woodward MC, et al. Among vitamin B12 deficient older people, high folate levels are associated with  worse cognitive function: combined data from three cohorts. J Alzheimers Dis 2014;39:661-8.
19.    van der Zwaluw NL, Dhonukshe-Rutten RA, van Wijngaarden JP, Brouwer-Brolsma EM, van de Rest O, In’t Veld PH, et al. Results of 2-year vitamin B treatment on cognitive performance: secondary data from an RCT. Neurology 2014;83:2158-66.
20.    Tangney CC, Tang Y, Evans DA, Morris MC. Biochemical indicators of vitamin B12 and folate insufficiency and cognitive decline. Neurology 2009;72:361-7.
21.    Brickman AM, Khan UA, Provenzano FA, Yeung LK, Suzuki W, Schroeter H, et al. Enhancing dentate gyrus function with dietary flavanols improves cognition in older adults. Nat Neurosci. 2014;17:1798-803.
22.    Lamport DJ, Pal D, Moutsiana C, Field DT, Williams CM, Spencer JP, et al.  The effect of flavanol-rich cocoa on cerebral perfusion in healthy older adults during conscious resting state: a placebo controlled, crossover, acute trial. Psychopharmacology (Berl). 2015 Jun 7. (Epub ahead of print) PubMed PMID: 26047963.
23.    Ide K, Yamada H, Takuma N, Park M, Wakamiya N, Nakase J, et al. Green tea consumption affects cognitive dysfunction in the elderly: a pilot study. Nutrients 2014;6:4032-42.
24.    Noguchi-Shinohara M, Yuki S, Dohmoto C, Ikeda Y, Samuraki M, Iwasa K, et al. Consumption of green tea, but not black tea or coffee, is associated with reduced risk of cognitive decline. PLoS One 2014;9:e96013.
25.    Suter PM, Haller J, Hany A, Vetter W. Diuretic use: a risk for subclinical thiamine deficiency in elderly patients. J Nutr Health Aging. 2000;4:69-71.
26.    Nazmi A, Weatherall M, Wilkins B, Robinson GM. Thiamin concentration in geriatric hospitalized patients using frusemide. J Nutr Gerontol Geriatr. 2014;33:47-54.
27.    Lengyel CO, Whiting SJ, Zello GA. Nutrient inadequacies among elderly residents of long-term care facilities. Can J Diet Pract Res. 2008;69:82-8.
28.    Koh F, Charlton K, Walton K, McMahon AT. Role of dietary protein and thiamine  intakes on cognitive function in healthy older people: a systematic review. Nutrients. 2015;7:2415-39.
29.    Wengreen H, Munger RG, Cutler A, Quach A, Bowles A, Corcoran C, et al. Prospective study of Dietary Approaches to Stop Hypertension- and Mediterranean-style dietary patterns and age-related cognitive  change: the Cache County Study on Memory, Health and Aging. Am J Clin Nutr 2013;98:1263-71.
30.    Tangney CC, Li H, Wang Y, Barnes L, Schneider JA, Bennett DA, et al. Relation of DASH- and Mediterranean-like dietary patterns to cognitive decline in older persons. Neurology 2014;83:1410-6.
31.    Pérez-López FR, Chedraui P, Haya J, Cuadros JL. Effects of the Mediterranean diet on longevity and age-related morbid conditions. Maturitas 2009;64:67-79.
32.    van de Rest O, Berendsen AA, Haveman-Nies A, de Groot LC. Dietary patterns, cognitive decline, and dementia: a systematic review. Adv Nutr. 2015;6:154-68.
33.    Singh B, Parsaik AK, Mielke MM, Erwin PJ, Knopman DS, Petersen RC, et al. Association of mediterranean diet with mild cognitive impairment and Alzheimer’s  disease: a systematic review and meta-analysis. J Alzheimers Dis 2014;39:271-82.
34.    Psaltopoulou T, Sergentanis TN, Panagiotakos DB, Sergentanis IN, Kosti R,  Scarmeas N. Mediterranean diet, stroke, cognitive impairment, and depression: A meta-analysis. Ann Neurol 2013;74:580-91.
35.    Chatzianagnostou K, Del Turco S, Pingitore A, Sabatino L, Vassalle C. The Mediterranean Lifestyle as a Non-Pharmacological and Natural Antioxidant for Healthy Aging. Antioxidants 2015;4: 719-36.
36.    Cooper C, Sommerlad A, Lyketsos CG, Livingston G. Modifiable predictors of dementia in mild cognitive impairment: a systematic review and meta-analysis. Am  J Psychiatry. 2015;172:323-34.
37.    Solfrizzi V, Panza F. Mediterranean diet and cognitive decline. A lesson from the whole-diet approach: what challenges lie ahead? J Alzheimers Dis. 2014;39:283-6.
38.    Monti MC, Margarucci L, Tosco A, Riccio R, Casapullo A. New insights on the interaction mechanism between tau protein and oleocanthal, an extra-virgin olive-oil bioactive component. Food Funct. 2011; 2:423-8.
39.    Bajerska J, Woźniewicz M, Suwalska A, Jeszka J. Eating patterns are associated with cognitive function in the elderly at risk of metabolic syndrome from rural areas. Eur Rev Med Pharmacol Sci 2014;18:3234-45.
40.    Ashby-Mitchell K, Peeters A, Anstey KJ. Role of dietary pattern analysis in determining cognitive status in elderly Australian adults. Nutrients 2015;7:1052-67.
41.    Azzini E, Polito A, Fumagalli A, Intorre F, Venneria E, Durazzo A, et al. 2011. Mediterranean  Diet Effect: an Italian picture. Nutr J 10:125. doi: 10.1186/1475-2891-10-125
42.    González-Guardia L, Yubero-Serrano EM, Delgado-Lista J, Perez-Martinez P, Garcia-Rios A, Marin C, et al. Effects of the Mediterranean diet supplemented with coenzyme q10 on metabolomic profiles in elderly men and women. J Gerontol A Biol  Sci Med Sci. 2015;70:78-84.
43.    Hernanz A, De la Fuente M, Navarro M, Frank A. Plasma aminothiol compounds, but not serum tumor necrosis factor receptor II and soluble receptor for advanced glycation end products, are related to the cognitive impairment in Alzheimer’s disease and mild cognitive impairment patients. Neuroimmunomodulation. 2007;14:163-7.
44.    Borroni B, Di Luca M, Padovani A. Predicting Alzheimer dementia in mild cognitive impairment patients. Are biomarkers useful? Eur J Pharmacol. 2006;545:73-80.
45.    Bermejo P, Martín-Aragón S, Benedí J, Susín C, Felici E, Gil P, et al. Peripheral levels of glutathione and protein oxidation as markers in the development of Alzheimer’s disease from Mild Cognitive Impairment. Free Radic Res. 2008;42:162-70.
46.    Mangialasche F, Polidori MC, Monastero R, Ercolani S, Camarda C, Cecchetti R, et al. Biomarkers of oxidative and nitrosative damage in Alzheimer’s disease  and mild cognitive impairment. Ageing Res Rev. 2009 ;8:285-305.
47.    Gu Y, Luchsinger JA, Stern Y, Scarmeas N. Mediterranean diet, inflammatory and metabolic biomarkers, and risk of Alzheimer’s disease. J Alzheimers Dis. 2010;22:483-9.
48.    García-Calzón S, Gea A, Razquin C, Corella D, Lamuela-Raventós RM, Martínez JA, et al. Longitudinal association of telomere length and obesity indices in an intervention study with a Mediterranean diet: the PREDIMED-NAVARRA trial. Int J Obes (Lond) 2014;38: 177–82.
49.    Boccardi V, Esposito A, Rizzo MR, Marfella R, Barbieri M, Paolisso G. Mediterranean diet, telomere maintenance and health status among elderly. PLoSOne. 2013;8:e62781.
50.    Sabatino L, Botto N, Borghini A, Turchi S, Andreassi MG. Development of a new  multiplex quantitative real-time PCR assay for the detection of the mtDNA(4977)deletion in coronary artery disease patients: a link with telomere shortening. Environ Mol Mutagen. 2013;54:299-307.
51.    Roberts RO, Boardman LA, Cha RH, Pankratz VS, Johnson RA, Druliner BR, et al. Short and long telomeres increase risk  of amnestic mild cognitive impairment. Mech Ageing Dev. 2014;141-142:64-9.
52.    Zhan Y, Song C, Karlsson R, Tillander A, Reynolds CA, Pedersen NL, et al. Telomere Length Shortening and Alzheimer Disease-A Mendelian Randomization Study. JAMA Neurol. 2015;72:1202-3.
53.    Rocha-González HI, Ambriz-Tututi M, Granados-Soto V. Resveratrol: a natural compound with pharmacological potential in neurodegenerative diseases. CNS Neurosci Ther. 2008;14:234-47.  
54.    Bastianetto S, Ménard C, Quirion R. Neuroprotective action of resveratrol. Biochim Biophys Acta. 2015;1852:1195-201.
55.    Braidy N, Jayasena T, Poljak A, Sachdev PS. Sirtuins in cognitive ageing and Alzheimer’s disease. Curr Opin Psychiatry. 2012;25:226-30.
56.    Shindler KS, Ventura E, Rex TS, Elliott P, Rostami A. SIRT1 activation confers neuroprotection in experimental optic neuritis. Invest Ophthalmol Vis Sci. 2007;48:3602-9.
57.    Hung CH, Chan SH, Chu PM, Tsai KL. Quercetin is a potent anti-atherosclerotic  compound by activation of SIRT1 signaling under oxLDL stimulation. Mol Nutr Food  Res Jul 23. doi: 10.1002/mnfr.201500144
58.    Bayram B, Ozcelik B, Grimm S, Roeder T, Schrader C, Ernst IM, et al. A diet rich in olive oil phenolics reduces oxidative stress in the heart of SAMP8 mice by induction of Nrf2-dependent gene expression. Rejuvenation Res 2012 15:71-81.
59.    Liang S, Steffen LM, Steffen BT, Guan W, Weir NL, Rich SS, et al. APOE genotype modifies the association between plasma omega-3 fatty acids and plasma lipids in the Multi-Ethnic Study of Atherosclerosis (MESA). Atherosclerosis 2013;228:181-7.
60.    Zhang L, Li M, Zhan L, Lu X, Liang L, Su B, et al. Plasma metabolomic profiling of patients with diabetes-associated cognitive decline. PLoS One. 2015;10:e0126952.
61.    Langie SA, Achterfeldt S, Gorniak JP, Halley-Hogg KJ, Oxley D, van Schooten FJ, et al. Maternal folate depletion and high-fat feeding from weaning affects DNA methylation and DNA repair in brain of adult offspring. FASEB J. 2013;27:3323-34.
62.    Bacalini MG, Friso S, Olivieri F, Pirazzini C, Giuliani C, Capri M, et al. Present and future of anti-ageing epigenetic diets. Mech Ageing Dev. 2014;136-137:101-15.
63.    Wang J, Freire D, Knable L, Zhao W, Gong B, Mazzola P, et al. Childhood and adolescent obesity and long-term cognitive consequences during aging. J Comp Neurol 2015;523:757-68.
64.    Santoro A, Pini E, Scurti M, Palmas G, Berendsen A, Brzozowska A, et al; NU-AGE Consortium. Combating inflammaging through a Mediterranean whole diet approach: the NU-AGE project’s conceptual framework and design. Mech Ageing Dev. 2014;136-7:3-13.

65.    UNESCO. Representative list of the intangible cultural heritage of humanity.
66.    Bach-Faig A, Berry EM, Lairon D, Reguant J, Trichopoulou A, Dernini S, et al. Mediterranean Diet Foundation Expert Group. Mediterranean diet pyramid today. Science and cultural updates. Public Health Nutr 2011;14:2274-84.
67.    Gallucci M, Antuono P, Ongaro F, Forloni PL, Albani D, Amici GP, et al. Physical activity, socialization and reading in the elderly over the age of seventy: what is the relation with cognitive decline? Evidence from «The Treviso  Longeva (TRELONG) study». Arch Gerontol Geriatr 2009;48:284-6.
68.    Holm L, Lund TB, Niva M. Eating practices and diet quality: a population study of four Nordic countries. Eur J Clin Nutr 2015;69:791-8.
69.    Keller H, Carrier N, Duizer L, Lengyel C, Slaughter S, Steele C. Making the most of mealtimes (M3): grounding mealtime interventions with a conceptual model. J Am Med Dir Assoc 2014;15:158-61.
70.    Whear R, Abbott R, Thompson-Coon J, Bethel A, Rogers M, Hemsley A, et al. Effectiveness of mealtime interventions on behavior symptoms of people with dementia living in care homes: a systematic review. J Am  Med Dir Assoc 2014;15:185-93.
71.    Yannakoulia M, Kontogianni M, Scarmeas N. Cognitive health and Mediterranean diet: just diet or lifestyle pattern? Ageing Res Rev 2015;20:74-8.
72.    Murphy T, Dias GP, Thuret S. Effects of diet on brain plasticity in animal and human studies: mind the gap. Neural Plast 2014;2014:563160.
73.    Gomez-Pinilla F, Gomez AG. The influence of dietary factors in central nervous system plasticity and injury recovery. PM R 2011;3:S111-6.